440 1 st St. NW, Ste. 520, Washington, D.C (202) , Fax (202) NPAinfo.org. March 19, 2018

Transcription

1 440 1 st St. NW, Ste. 520, Washington, D.C (202) , Fax (202) NPAinfo.org Dockets Management Staff (HFA-305) Division of Dockets Management U.S. Food and Drug Administration 5630 Fishers Lane, Rm Rockville, MD RE: Docket Number FDA-2017-N-0763 for Food Labeling: Health Claims; Soy Protein and Coronary Heart Disease; Extension of Comment Period. 83 Federal Register (Publication Date: January 17, 2018). Dear FDA Desk Officer, The Natural Products Association (NPA) appreciates the opportunity to respond to the U.S. Food and Drug Administration s (FDAs) notification of request for public comment on Food Labeling: Health Claims; Soy Protein and Coronary Heart Disease. NPA is submitting this document as a formal comment to the Center for Safety and Applied Nutrition s (CFSAN s) request for information seeking input from the public on its proposal to revoke a regulation authorizing the use of health claims on the relationship between soy protein and coronary heart disease on the label or in the labeling of foods. NPA has concerns over FDA s interest to revoke the heart health claim given the number of studies and meta analyses performed over the years that were used as the competent and reliable scientific evidence to support this claim on labels and in labeling to inform consumers. Many NPA members are also classified as small business manufacturers, distributors, re-packers, and retailers who would be affected by this proposal to revoke the heart health claim in the absence of performing an economic impact analysis.

2 Page 2 Executive Summary The totality of evidence indicates a clear benefit in heart health as a result of the positive association between soy protein consumption and LDL cholesterol reduction. FDA should not eliminate the soy protein heart health claim or convert it to a qualified health claim. FDA does not have any material fact to compel industry to change labeling by eliminating the soy protein heart health claim or converting it to a qualified health claim. All material facts support keeping the heart health claim and lowering the threshold to 15 g soy protein. NPA has performed its own meta-analysis, reporting an overall effect size reduction in LDL cholesterol of ±.024 mmol/l NPA s meta-analysis demonstrating a positive association between soy protein consumption and lowering of LDL cholesterol is similar in magnitude to others performed after NPA encourages FDA to replace 25 g in the current heart health claim with 15 g as the new threshold in soy protein a newer meta-analysis from 2011 supports this serving level to consistently lower LDL cholesterol and therefore benefit heart health. Soy protein lowers LDL cholesterol through an intrinsic or direct effect and an extrinsic effect by consumers replacing animal protein sources with plant-based soy protein sources. The combined intrinsic and extrinsic effect of soy to lower LDL cholesterol is 10.3%, which is on par with the heart health claim for plant sterols and stanol esters to reduce LDL cholesterol by 10%. FDA has failed to perform its own meta-analysis and Forest plot out of government transparency on the totality of study evidence to make a determination on the overall effect of soy protein on LDL cholesterol. Without such data, it is unclear how FDA, a data-drive public health agency, can propose to eliminate the health claim.

3 Page 3 FDA fails to meet the four-part Central Hudson test to compel a labeling change in its proposal to eliminate the soy heart health claim in the absence of any scientific justification; FDA s proposal to do so is misleading on its face. Background NPA was founded in 1936 to promote and protect the unique values and shared interests of retailers and suppliers of natural nutritional foods, dietary supplement, and other natural consumer products. NPA is a non-profit 501(c)(6) association whose mission is to advocate for the rights of consumers to have access to products that will maintain and improve their health, and for the rights of retailers and suppliers to sell these natural products. We are the oldest and largest trade association in the natural products industry representing over 1,000 members accounting for almost 10,000 retail, manufacturing, wholesale, and distribution locations of natural products, including conventional foods, dietary supplements, medical foods, and health/beauty aids. Therefore, the NPA and its member companies have a significant interest in this proposal to revoke the soy protein hearth health claim. The natural products industry is important not only to the health and wellbeing of millions of Americans but also to the health of the United States economy. In 2011, according to the Nutrition Business Journal, sales of natural products grew 8.4 percent to $126 billion Estimates of the natural, organic and healthy consumer products are predicted to expand with a compound average growth rate of nearly 9 percent over the next 2 years to $252B by This is more than four times the projected growth rate of mainstream consumer packaged goods. Due to a widespread lifestyle shift, caused by a multitude of factors including greater environmental awareness, millennial purchasing trends, and an improving health and wellness education, many of today s consumers in the U.S. are demanding greater transparency in their purchases. Consumers seek to know where their products come from as well as their exact composition.

4 Page 4 Introduction The FDA is proposing to revoke its regulation authorizing the use of health claims on the relationship between soy protein and coronary heart disease on the label or in the labeling of foods. 1 FDA is taking this action based on their supposed review of the totality of publicly available scientific evidence currently available and their tentative conclusion that such evidence does not support their previous determination that there is significant scientific agreement (SSA) among qualified experts for a health claim regarding the relationship between soy protein and reduced risk of coronary heart disease. In 1999, FDA authorized a health claim about the relationship between soy protein and a reduction in the risk of coronary heart disease (CHD). In the Federal Register of December 21, 2007, FDA announced an intention to reevaluate the scientific evidence for this health claim and provided the opportunity for public comment. 2 FDA explained that they were reevaluating the scientific basis for the soy protein and CHD health claim because newer studies yielded varied and inconsistent findings ranging from beneficial effect to no statistically significant effect from one trial to another. FDA claimed that the results of these studies called into question the conclusions drawn from their prior review, which had served as the basis for authorizing the soy protein and reduced risk of CHD health claim. Today FDA has decided to revoke the soy protein and CHD claim based upon the totality of the publicly available scientific evidence that, although they admit that some evidence does suggest a relationship between soy protein intake and reduced risk of CHD, the totality of the evidence is inconsistent and not conclusive. FDA has taken this stance with their determination that the strength of the totality of the publicly available data does not meet the SSA standard for a relationship between soy protein intake and CHD risk. 3,4 However, NPA would like to delve into the scientific support with greater scrutiny here. NPA believes the totality of the evidence favors a continuation of (21 CFR ) FR Federal Food, Drug, and Cosmetic Act 403(r)(3) (21 U.S.C. 343(r)(3)) 4 FDA 2009 Health Claim Guidance Evidence-Based Review System for the Scientific Evaluation of Health Claims

5 Page 5 use on labels and in labeling of the association of soy protein and reduction in CHD risk. NPA has performed a meta-analysis on soy protein clinical trial data collected after 1995 to determine the overall effect size of soy s intrinsic effect. What is Soy Protein? Soy protein, according to FDA s description in the Federal Register authorizing claims for a reduction in CHD risk, 5 is an edible component of the soybean, Glycine max. Soy protein is produced from raw whole soybeans by a multistep process that removes the lipid and indigestible components to concentrate the protein and increase its availability. Depending on the particular steps used during the processing, soy protein ingredients may take the form of isolated soy protein, soy protein concentrate, or soy flour. Each ingredient may be further processed into texturized soy protein or texturized vegetable protein used in the manufacture of meat and poultry analogues by thermoplastic extrusion or steam texturization to impart structure and shape. In addition to protein, these soy protein ingredients contain other naturally occurring soy constituents, such as isoflavones, fiber, and saponins. The specific processing steps that are used determine the extent of retention of naturally occurring components in the final product. Another difference in soy concentrates and isolates is whether the manufacturer used water or ethanol washing to concentrate the protein as ethanol washing removes most of the isoflavones and saponins. Many studies involving soy concentrates and isolates do not report on the processing conditions and composition information. The constituent content varies depending on the manufacturing process chosen. Texturized vegetable protein is usually made from a combination of soy flour and soy protein concentrate. Soy flour is typically comprised of 50% protein and 5% fiber; soy concentrate is 65-70% protein and sparse fiber; isolated soy protein contains a majority (typically 90%) protein with no fiber content. 6 Soy protein contains all of the essential amino acids to support human 5 U.S. Food and Drug Administration Food labeling, health claims, soy protein, and coronary heart disease. Fed Reg 57: Protein Technologies International Soy Protein and Health: Discovering a Role for Soy Protein in the Fight Against Coronary Heart Disease. Houston, TX: Marimac Communications.

6 Page 6 life and is therefore recognized as a complete protein. Components of soy protein have been implicated in its hypocholesterolemic benefit including trypsin inhibitors, phytic acid, saponins, isoflavones, and fiber. Soy protein is also consumed as a component of traditional fermented and nonfermented soy foods, such as tofu, tempeh, and miso, as well as whole soybeans, soynuts, soymilk, soy yogurt, and soy cheese. These products contain variable amounts of soy protein and other naturally occurring soy constituents depending on the specific technologies used in these products. Soy protein ingredients and soy protein-containing foods may partially replace or be used in addition to animal or other vegetable protein sources in the human diet. Clearly, not all soy proteins out in the market are born the same. These differences may underlie some of the inconsistencies in some of the later studies designed to further investigate components of soy and risk reduction of cardiovascular disease. Evidence for Soy Protein and Reduction in Risk in Cardiovascular Disease (Past 1999) Cardiovascular disease is the leading cause of death in the United States. 7 Diet was demonstrated anecdotally to have a profound impact on several modifiable risk factors for heart disease: hypercholesterolemia, elevated low density lipoprotein (LDL or bad ) cholesterol, low HDL (high density or good ) cholesterol, hypertriglyceridemia, hypertension, obesity, and diabetes. A low-saturated fat, low cholesterol diet, recommended among heart experts, does help to lower the risk for cardiovascular disease; 8,9 however, other dietary factors may also offer additional heart health benefits. Considerable evidence mounted in the 1990s that consumption of soy protein in place of animal protein lowers blood cholesterol levels and may provide other cardiovascular 7 CDC, NCHS Underlying cause of death on CDC WONDER Online Database, released Data are from the Multiple Cause of Death Files, , as compared from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative Program. Accessed Feb 7, Krauss R.M., Deckelbaum R.J., Ernst N., et al Dietary guidelines for healthy American adults. Circulation 94: Yu-Poth S., Zhao G., Etherton T., et al Effects of the National Cholesterol Education Program s Step I and Step II dietary intervention programs on cardiovascular disease risk factors: a meta-analysis. Am J Clin Nutr 69:

7 Page 7 benefits. Epidemiological evidence collected in Asian populations, known to consume greater quantities of soy foods as part of their dietary staple, indicates a lower predilection and incidence of cardiovascular disease in this population than in those who consume a typical Western diet. 10 Soy protein is consumed in Japan at serving levels as great as 55 g/day, 11 compared with less than 5 g/day in the United States. 12 In 1998, disease deaths per 100,000 people between the ages of 35 and 74 years included 401 for U.S. men, 201 for Japanese men, 197 for U.S. women, and 99 for Japanese women. 13 There are many differences in dietary patterns and lifestyle factors that could also help account for these differences in disease patterns among industrialized nations, but diet is considered a major indicator. During the 1990s, the American Heart Association s (AHA s) Dietary Guidelines for Healthy American Adults 14 supported evidence that when soy protein was substituted for animal protein, total and LDL cholesterol could be reduced, the findings were inconclusive. Although the AHA Nutrition Committee concluded that the use of soy foods was consistent with the AHA Dietary Guidelines, no further recommendation was ever made to include soy protein in the U.S. diet. Instead more research on the mechanism of action explaining the beneficial effects of soy protein and related phytochemicals on blood lipids was recommended. It appeared that the understanding of the mechanism of action was more important than the actual clear benefit attained from replacing animal protein with soy protein. Landmark Meta-Analysis Characterizing Soy s Benefit on Heart Health The landscape of soy protein changed in A landmark meta-analysis by Anderson et al. of 38 controlled clinical studies was published in the New England Journal of Medicine and 10 Beaglehole R International trends in coronary heart disease mortality, morbidity, and risk factors. Epidemiol Rev 12: Nagata C., Takatsuka N., Kurisu Y., et al Decreased serum total cholesterol concentration is associated with high intake of soy products in Japanese men and women. J Nutr 128: Messina M., Messina V., and Setchell K The Simple Soybean and Your Health. Garden City Park, NY: Avery Publishing Group. 13 American Heart Association Heart and Stroke Statistical Update. Dallas, TX: American Heart Association; Krauss R.M., Deckelbaum R.J., Ernst N., et al Dietary guidelines for healthy American adults. Circulation 94:

8 Page 8 concluded that substituting soy protein for animal protein significantly lowered total cholesterol, LDL cholesterol, and triglycerides without affecting HDL cholesterol. These effects were also noted to be greater in subjects with higher baseline cholesterol values. Daily soy protein consumption resulted in a 9.3% decrease in total serum cholesterol, a 12.9% decrease in LDL cholesterol, and a 10.5% decrease in triglycerides. 15 The direct cholesterol-lowering effect of soy (or so called intrinsic effect of soy) was in addition to the effect seen with a diet low in saturated fat and cholesterol (extrinsic effect of replacing animal protein with soy protein from plant sources). Studies included in the New England Journal of Medicine manuscript used soy protein in the form of either texturized vegetable protein or isolated soy protein. No difference in efficacy was noted between these sources of soy protein in this metaanalysis, although the compositions of these soy products were quite different. As will be discussed later in these comments, the studies analyzed by Anderson et al. prior to 1996 were probably not normalized for macronutrients profile. Therefore, this meta-analysis included studies with both an intrinsic and extrinsic cholesterol lowering effect of soy. The intrinsic v. extrinsic effect of soy will be discussed later on in these comments. Both effects are present when consumers switch from eating animal protein to soy protein and therefore the effects additive to reduce coronary heart disease risk. Effect of Soy Protein in Adults with Normal Cholesterol Levels Consumption of soy protein does not have a hypocholesterolemic effect in normal, healthy adults having cholesterol levels within the normal range or low cholesterol. Therefore, there is no concern for soy reducing cholesterol to a hypocholesterolemic range. In a study of 12 adults wth a mean total cholesterol level of 145 mg/dl at baseline, the incorporation of 66 to 80 gm soy protein (meat replaced by soy analogues and milk replaced by soy beverage) 15 Anderson J.W., Johnstone B.M., Cook-Newell M.E Meta-analysis of the effects of soy protein intake on serum lipids. N Engl J Med 333:

9 Page 9 resulted in no significant changes in serum lipids. 16 Sacks et al. found no significant changes in serum lipids in 13 strict vegetarians whose baseline cholesterol was 129 mg/dl. 17 A similar study by Wong et al. found no significant change in serum lipids in 13 normocholesterolemic men 20 to 50 years of age (mean baseline total cholesterol, 169 mg/dl) who consumed 50 gm soy protein in addition to a diet low in saturated fat and cholesterol. 18 In the New England Journal of Medicine meta-analysis on the effect of soy protein on serum cholesterol levels by Anderson et al., no significant effect of soy protein was found for those with a total cholesterol < 200 mg/dl. 15 Soy Protein Studies in Adults with Elevated Cholesterol Levels While there are some conflicting results in studies in adults with elevated serum cholesterol levels, most studies report total and LDL cholesterol reductions after the addition of soy protein to a diet that is low in saturated fat and cholesterol. In a study of postmenopausal women on a diet low in saturated fat and cholesterol and consumption of 40 gm/d of soy protein with either 56 or 90 mg of isoflavones daily or casein for 6 months, both soy groups had significantly better blood lipid profiles (average change from baseline, 8.2% decrease in non- HDL cholesterol and a 4.4% increase in HDL cholesterol) than the casein group. 19 Furthermore, no differences in lipids were seen between the two isoflavone levels. Another study indicated that HDL significantly increased 7% from baseline with consumption of 32 gm soy protein as soymilk in both women and men with hypercholesterolemia. 20 Similar studies on the effect of soy protein as part of a hypocaloric diet found significantly lower total and LDL cholesterol 16 Giovannetti P.M., Carroll K.K., Wolfe B.M Constancy of fasting serum cholesterol of healthy young women upon substitution of isolated soy protein for meat and dairy protein in medium and low fat diets. Nutr Res 6: Sacks F.M., Breslow J.L., Wood P.G., et al Lack of an effect of dairy protein (casein) and soy protein on plasma cholesterol of strict vegetarians: an experiment and a critical review. J Lipid Res 24: Wong W.W., Smith E.O., Stuff J.E., et al Cholesterol-lowering effect of soy protein in normocholesterolemic and hypercholesterolemic men. Am J Clin Nutr 1385S-1389S. 19 Baum J.A., Teng H., Erdman J.W. Jr., et al Long-term intake of soy protein improves blood lipid profiles and increases mononuclear cell low-density-lipoprotein receptor messenger RNA in hypercholesterolemic, postmenopausal women. Am J Clin Nutr 68: Kurowska E.M., Jordan J., Spence J.D., et al Effects of substituting dietary soybean protein and oil for milk protein and fat in subjects with hypercholesterolemia. Clin Invest Med 20:

10 Page 10 compared with a conventional hypocaloric diet. 21,22 Another study from Crouse et al. reported reduction of 4% and 6% in total and LDL cholesterol, respectively, in hypercholesterolemic individuals consuming 25 gm soy protein with isoflavones as part of a diet low in saturated fat and cholesterol. 23 A 9-week human study comparing the effects of soy protein (25 g/d) containing varying levels of isoflavones with those of casein found that consumption of the highest isoflavone level (62 mg/d) resulted in significantly lower total and LDL-cholesterol values than those of the casein group. Subjects with the highest LDL-cholesterol levels (top 50%) also experienced significant decreases in total and LDL cholesterol with37 mg/d of isoflavones. Those consuming soy protein at 27 mg/d or less showed no significant cholesterol-lowering effect. 23 Evidence for Soy Protein and Reduction in Risk in Cardiovascular Disease (1999 Present) The Nutrition Committee of the American Heart Association assessed 22 randomized clinical trials published since The vast majority of studies (19/22) were conducted with hyperlipidemic subjects, which is the subpopulation where an effect would be predicted to be observed as the previous decade showed no statistically significant decrement in LDL in normolipidemic volunteers These studies showed that following a mean consumption of 50 g/day ( g range) of isolated soy protein containing isoflavones, LDL cholesterol levels were lowered by 3% in comparison with milk or other proteins. In only a small subgroup of these studies was the effect of soy protein on LDL cholesterol and other lipid risk factors shown to be insignificant. A separate post-menopausal women study by Welty et al. in 2007 involving consumption of 25 g soy protein and 101 mg of aglycone isoflavone reduced LDL cholesterol 21 Jenkins D.J.A., Wolever T.M.S., Spiller G., et al Hypocholesterolemic effect of vegetable protein in a hypocaloric diet. Atherosclerosis 78: Bossello O., Cominacini L., Zocca I., et al. Short-and long-term effects of hypocaloric diets containing proteins of different sources on plasma lipids and apoproteins of obese subjects. Ann Nutr Metab 32: Crouse J.R III, Morgan T., Terry J.G., et al A randomized trial comparing the effect of casein with that of soy protein containing varying amounts of isoflavones on plasma concentrations of lipids and lipoproteins. Arch Intern Med 159: Sacks F.M., Lichtenstein A., Van Horn L., Harris W., Kris-Etherton P., and Winston, M Soy protein, isoflavones, and cardiovascular health. An American Heart Association Science Advisory for Professionals from the Nutrition Committee. Circulation 113:

11 Page 11 levels by 11%, systolic blood pressure, and diastolic blood pressure by 11%, 9.9% and 6.8%, respectively, in hypertensive women. 25 A 2009 study of slightly or mildly hypercholesterolemic subjects including men and post-menopausal women older than 45 years demonstrated a 5.5 ± 1.6% decrease in total and LDL cholesterol. Some researchers may point to the meta-analysis by Yeung and Yu 26 in 2003 to conclude no effect on LDL cholesterol. However, it should be noted that while this manuscript might have been peer-reviewed, this is a pay-to-play, open access journal, and the scientific rigors of which can and should be questioned. The meta-analysis concludes there is no effect (effect outcome of 0.01 in favor of control) after including 17 studies. 27,28,29,30,31,32,33,34,35,36 On closer inspection, one notices that their conclusion is based upon isoflavone tablets, isolate or mixture, with up to 150 mg/d. Upon an even closer inspection, the studies lumped together were vastly different. 25 Welty F.K., Lee K.S., Lew N.S., and Zhou J.R Effect of soy nuts on blood pressure and lipid levels in hypertensive, prehypertensive, and normotensive postmenopausal women. Arch Intern Med 167: Yeung J. and Yu T.F Effects of isoflavones (soy phyto-estrogens) on serum lipids: a meta-analysis of randomized controlled trials. Nutr J 2: 15 (open access journal). 27 Clifton-Bligh P.B., Baber R.J., Fulcher G.R., Nery M.L., and Moreton T The effect of isoflavones extracted from red clover (Rimostil) on lipid and bone metabolism. Menopause 8: Dewell A., Hollenbeck C.B., and Bruce B The effects of soy-derived phytoestrogens on serum lipids and lipoproteins in moderately hypercholesterolemic postmenopausal women. J Clin Endocrinol Metab 87: Hale G., Paul-Labrador M., Dwyer J.H., and Merz C.N.B Isoflavone supplementation and endothelial function in menopausal women. Clin Endocrinol 56: Hodgson J.M., Puddy I.B., Beilin L.J., Mori T.A., and Croft K.D Supplementation wih isoflavonoid phytoestrogens does not alter serum lipid concentrations: a randomized controlled trial in humans. J Nutr 128: Howes J.B., Sullivan D., Lai N., Nestel P., Pomeroy S., West L., Eden J.A., and Howes L.G The effects of dietary supplementation with isoflavones from red clover on the lipoprotein protein profiles of post menopausal women with mild to moderate hypercholesterolaemia. Atherosclerosis 152: Nestel P.J., Yamashita T., Sasahara T., Pomeroy S., and Dart A Soy isoflavones improve systemic arterial compliance but not plasma lipids in menopausal and perimenopausal women. Arterioscler THromb Vasc Bull 17: Nestel P.J., Pomeroy S., Kay S., Komesaroff P., Behrsing J., Cameron J.D., and West L Isoflavones from red clover improve systemic arterial compliance but not plasma lipids in menopausal women. J Clin Endocrinol Metab 84: Samman S., Wall P.M.L., Chan G.S.M., Smith S.J., and Petocz P The effect of supplementation with isoflavones on plasma lipids and oxidiability of low density lipoprotein in premenopausal women. Atherosclerosis 147: Simons L.A., von Konigsmark M., Simons J., and Celermajer D.S Phytoestrogens do not influence lipoprotein levels or endothelial function in healthy, postmenopausal women. Am J Cardiol 85: Squadrito F., Altavilla D., Morabito N., Crisafulli A., D Anna R, Corrado F., Ruggeri P., Campo G.M., Calapai G., Caputi A.P., and Squadrito G The effect of the phytoestrogen genistein on plasma nitric oxide concentrations, endothelin-1 levels and endothelium dependent vasodilation in postmenopausal women. Atherosclerosis 163:

12 Page 12 Some studies included soy burgers, isolated soy protein, soy protein capsules, and soy biscuits while others included isoflavone tablets alone. Some studies included males or females on while others included post- or peri-menopausal women or elderly men alone. The study was vastly heterogenous in both population and design to be useful in making any aggregate assessments from any final meta-analysis effect outcome generated. After stating their conclusions that isoflavone tablets did not lower total or LDL-cholesterol, the authors included a critical disclaimer that the studies including isolated soy protein, soy diets or soy protein capsules were inconclusive due to inadequate sample size, heterogeneity [within those clinical trials] and presence of potentially uncontrolled cofounders [not identified by the authors]. NPA has performed its own meta-analysis on 54 clinical trials (see Appendix; Table 1). NPA s analysis among 54 clinical trials on LDL cholesterol reports an overall effect size reduction in LDL cholesterol of ±.024 mmol/l by consumption of soy protein (10 g- 170 g). This corresponds to an approximately 5.6% reduction overall, which is very similar to other metaanalyses on the intrinsic effect of soy protein, on LDL cholesterol. NPA used a fixed effect model in its analysis (Appendix; Table 2). NPA did NOT incorporate any studies performed prior to 1996 as in the Anderson et al. meta-analysis because this meta-analysis already demonstrated the combined intrinsic and extrinsic effect of soy on heart health. NPA wanted to determine the scientific agreement among soy protein on LDL cholesterol in clinical trials performed after 1996, when macronutrient profiles were controlled. NPA performed a gold standard Forest plot on the association of soy protein and LDL cholesterol reduction (see Appendix; Fig.1). NPA points out that FDA has failed to perform its own meta-analysis and Forest plot on the totality of the evidence to make a determination on the overall effect of soy protein on LDL cholesterol. Without such data, how can FDA propose to eliminate the soy protein heart health claim? The effect of soy is clear as the majority of studies report a reduction in LDL cholesterol and therefore scattered to the left of zero on the abscissa. These clinical trials capture the intrinsic effects of soy protein. The study of Yeung and Yu was not included in our analysis because of issues described above in sample size, heterogeneity in population or heterogeneity in the actual test article given (isoflavones tablets vs. actual soy proteins). NPA used studies where

13 Page 13 there was sufficient data reported in the study to utilize in the analysis. NPA s analysis did include studies showing no effect by soy protein to consider the scientific agreement among all studies in totality. Any studies favoring control in the Forest plot and therefore showing no effect of soy protein on LDL cholesterol are indicated to the right of zero on the abscissa. The totality of evidence indicate a clear benefit of soy protein on heart health as evidenced from the overall effect size in the analysis (see Appendix; Table 1 and Fig. 1). A newer, updated meta-analysis from J.W. Anderson and Bush in 2011 corroborated findings from their earlier 1995 New England Journal of Medicine meta-analysis. 37 Anderson s 2011 meta-analysis incorporated 43 clinical trials (20 parallel-design and 23 crossover studies) related to soy protein intake and serum lipoprotein changes between 1996 and They concluded findings consistent from their 1995 meta-analysis that soy protein consumption with a median of 30 g/d was associated with a significant improvement in lipoprotein risk factors for coronary heart disease. They also found that parallel RCTs were associated with significantly greater improvements in serum LDL-cholesterol compared to crossover studies. They concluded that regular consumption of 1 to 2 servings of soy protein daily (15 to 30 g) has a significant favorable impact on serum lipoprotein risk factors for coronary heart disease. Effect of Isolated Soy Components Several studies have investigated the specific roles of isoflavones, soy protein and their interaction to induce a hypocholesterolemic effect. While discrepancies remain in relation to the actual lipid-lowering effects of soy constituents 39, several studies have concluded that consumption of isoflavone alone does not reduce lipids. 40,41,42 It should be noted that while the 37 Anderson J.W. and Bush H.M Soy protein effects on serum lipoproteins: a quality assessment and metaanalysis of randomized, controlled studies. J Am Coll Nutr 30: The meta-analysis by J.W. Anderson and colleagues in 1995 previously included a comprehensive search of clinical trials conducted in the public domain up through Lichtenstein A Soy protein, isoflavones and cardiovascular disease risk. J Nutr 128: Yeung J. and Yu T.F Effects of isoflavones (soy phyto-estrogens) on serum lipids: a meta-analysis of randomized controlled trials. Nutr J 2: 15 (open access journal).

14 Page 14 Nestel et al. study from 1997 did not show any effect on lowering LDL cholesterol, it did show improvement in systemic arterial compliance and therefore an impact on heart health. While isoflavones by themselves may not be the silver bullet to lower LDL cholesterol, they are critical and necessary in combination with other soy constituents to result in a dose-dependent lowering of LDL cholesterol when combined with soy protein. 43,44,45 While the reason why a combination is necessary, there is general agreement that a soy protein matrix containing both isoflavones AND intact soy protein appears to be more beneficial than either component alone. 46,47 A Case for the Consideration of Micro Gut Flora Gut microflora plays an essential role in isoflavone bioavailability because of the necessity of bacterial enzymes required for isoflavone absorption and metabolism. Studies have shown that plasma and urinary concentrations of isoflavones can vary widely between individuals after eating soy proteins. 48,49 A study from Larkin et al. in 2009 seemed to support a 41 Nestel P., Yamashita T., Sasahara T., Pomeroy S., Dart A., Komesaroff P. et al Soy isoflavones improve systemic arterial compliance but not plasma lipids in menopausal and perimenopausal women. Arterioscler Thromb Vasc Biol 17: Hodgson J.M., Puddy I.B., Beilin L.J., Mori T.A., and Croft K.D Supplementation wih isoflavonoid phytoestrogens does not alter serum lipid concentrations: a randomized controlled trial in humans. J Nutr 128: Crouse J., Morgan T., Terry J., Ellis J., Vitolins M., and Burke G A randomized trial comparing the effect of casein with that of soy protein containing varying amounts of isoflavones on plasma concentration of lipids and lipoproteins. Arch Intern Med 59: Merz-Demlow B., Duncan A., Wangen K., Xu X., Carr T., Phipps W. et al Soy isoflavones improve plasma lipids in normocholesterolemic, premenopausal women. Am J Clin Nutri 71: Wangen K., Duncan A., Xu X., and Kurzer M Soy isoflavones improve plasma lipids in normocholesterolemic and mildly hypercholesterolemic postmenopausal women. Am J Clin Nutr 73: Potter S Soy protein and cardiovascular disease: the impact of bioactive components in soy. Nutr Rev 56: Steinberg F., Guthrie N., Villablanca A., Kumar K., Murray M Soy protein with isoflavones has favorable effects on endothelial function that are independent of lipid and antioxidant effects in healthy postmenopausal women. AM J Clin Nutr 78: Winter J.and Bokkenheuser V Bacterial metabolism of natural and synthetic sex hormones undergoing enterohepatic circulation. J Steroid Biochem 27: Rowland I., Wiseman H., Sanders T., Adlercreutz H., and Bowey E Metabolism of oestrogens and phytoestrogens: role of the gut microflora. Biochem Soc Trans 27:

15 Page 15 role from soy protein independent of isoflavone bioavailability. 50 This study suggested that dietary combination of soy with either a probiotic or a prebiotic resulted in significant lowering of total (5.5 ± 1.5%) and LDL cholesterol (7.3± 2.2%). If isoflavones are indeed necessary for the hypocholesterolemic effect of soy, differences between individuals in gut microflora and isoflavone bioavailability may underline any variability observed in variable lipid responses among clinical trials. Soy Threshold Required to Reduce Cholesterol Researchers have also investigated as to whether there is a floor of dietary soy protein required to reduce cholesterol. A dose-response study in hypercholesterolemic men on a National Cholesterol Education Program (NCEP) Step I diet (low saturated fat and cholesterol) involving g/day of soy protein in comparison with casein, demonstrated significantly greater reductions in non-hdl cholesterol (1.5% to 4.5% range) than the casein control group for all doses. The cholesterol lower effect by soy was also dose-dependent. 51 In response to these studies, FDA published their final ruling on the health claim for soy protein and reduced risk of heart disease as a result of cholesterol reduction, setting their threshold at 25 g/day of soy protein when combined with a diet low in saturated fat and cholesterol. 52 The more recent meta-analysis from Anderson and Bush in 2011 concluded that regular consumption of 1 to 2 servings of soy protein daily (15 to 30 g/day) has a significant favorable impact on serum lipoprotein risk factors for coronary heart disease. This meta-analysis would seem to support that a threshold of 15 g of soy protein is also sufficient to lower cholesterol. 53 Based upon the conclusions of this newer meta-analysis by Anderson and Bush, NPA encourages FDA to replace 50 Larkin T.A., Astheimer L.B., and Price W.E Dietary combination of soy with a probiotic or prebiotic food significantly reduces total and LDL cholesterol in mildly hypercholesterolaemic subjects. Eur J Clin Nutr 63: Teixeira S., Potter S.M., Weigel R., et al Effects of feeding 4 levels of soy protein for 3 and 6 weeks on blood lipids and apolipoproteins in moderately hypercholesterolemic men. Am J Clin Nutr 71: U.S. Food and Drug Administration Food labeling, health claims, soy protein, and coronary heart disease. Fed Reg. 57: Anderson J.W. and Bush H.M Soy protein effects on serum lipoproteins: a quality assessment and metaanalysis of randomized, controlled studies. J Am Coll Nutr 30:

16 Page g in the current heart health claim with 15 g as the new threshold for soy protein to support a lowering of LDL cholesterol and therefore a heart health claim. Mechanism of Action by Soy Protein Trypsin inhibitors are ubiquitous in foods. All soy products typically undergo thermal processing, which negates most of the activity of trypsin inhibitors. Small amounts of the heatstable Bowman-Birk inhibitor may exert a hypocholesterolemic effect by increasing the secretion of cholecystokinin, which stimulates bile acid synthesis from cholesterol, leading to reductions in cholesterol. Animal studies however failed to demonstrate any cholesterol lowering effect when trypsin inhibitor was added alone to the diet. 54 Phytic acid and copper are found in all nonfermented soy protein products. Phytic acid, which is heat stable, chelates zinc strongly in the intestinal lumen, thus decreasing its absorption. 55 A copper deficiency or a high free zinc-to-copper ratio also results in a rise in blood cholesterol. 56 Soy may therefore be exerting its cholesterol lowering effect through chelation by decreasing the ratio of zinc to copper. Saponins are heat-stable compounds present in all soy protein products except those extracted with alcohol. These compounds may contribute to soy s cholesterol reducing effect by increasing bile excretion. 57 Intrinsic and Extrinsic Effect The discrepancy between the large LDL cholesterol reductions reported in older soy protein studies before 1996 compared with smaller reductions in more recent years requires further exploration. Consensus agreement is that this discrepancy is due to intrinsic v. extrinsic effects of soy. Studies have investigated whether there is a separate physiologically meaning LDL reduction resulting from the displacement from the diet of saturated fat and cholesterol- 54 Roy D.M., Schneeman B.O Effect of soy protein, casein and trypsin inhibitor on cholesterol, bile acids and pancreatic enzymes in mice. J Nutr 111: Zhou J.R., Erdman J.W. Jr Phytic acid in health and disease. Crit Rev Food Sci Nutr 35: Klevay L.M Coronary heart disease: the zinc/copper hypothesis. Am J Clin Nutr 28: Sidhu G.S., Oakenfull D.G A mechanism for the hypocholesterolemic activity of saponins. Br J Nutr 55:

17 Page 17 rich foods by soy food consumption. This is the so called extrinsic soy effect. Soy protein consumption results in significant reductions in LDL cholesterol as demonstrated by the New England Journal of Medicine study by Anderson et al. in The intrinsic soy effect has been demonstrated over and over in intervention studies and meta-analyses that demonstrated equivalent macronutrient profiles, but contributions from the extrinsic soy effect had been harder to quantify. The data from the 1995 Anderson et al. study suggested that soycontaining foods will have an even greater effect on LDL cholesterol reduction by displacing foods high in saturated fat and cholesterol. One study answered this question by examining the intrinsic cholesterol-lowering effect of soy by performing a meta-analysis on the AHA soy advisory data to determine the LDL cholesterol reduction in soy studies where test and control diets had the same macronutrient profiles. 58 A sensitivity analysis on the original Anderson et al. study in 1995 was performed for comparison. In addition, the Third National Health and Nutrition Examination Survey (NHANES III) population survey data was used to estimate the potential LDL cholesterol reduction from substitution of foods containing more saturated fat and cholesterol-rich animal products (diets built upon the 50 th percentile intake data in NHANES III for total and saturated fat) with soy containing foods (extrinsic effect). Meta-analysis of the AHA Soy Advisory data gave a statistically significant LDL cholesterol reduction of 0.17 mmol/l (n=22; P<0.0001), corresponding to a 4.3% reduction in LDL cholesterol for soy. This was confirmed in 11 studies reporting balanced macronutrient profiles across all studies. The isolated extrinsic effect of soy was due to displacement of saturated fats and cholesterol from animal food diets and calculated to be between % when displacing with soy (13-58 g/day range) using NHANES III population survey data. Therefore, a combined intrinsic and extrinsic effect of soy protein foods ranges anywhere between 7.9 to 10.3%. Therefore, soy remains one of the few food components that has 58 Jenkins D.J.A., Mirrahimi A., Srichaikul K., Berryman C.E., Wang L., Carleton A., Abdulnour S, Sievenpiper J.L., Kendall C.W.C., and Kris-Etherton P.M Soy protein reduces serum cholesterol by both intrinsic and food displacement mechanisms. J Nutr (Supplement: Soy Summit Exploration of the Nutrition and Health Effects of Whole Soy)140: 2302S-2311S.

18 Page 18 consistently been shown to reduce serum cholesterol by greater than 4% when added to the diet. The intrinsic-extrinsic soy hypothesis offers an explanation for why earlier studies showed such a large reduction and recent studies showed less of a reduction. The intrinsic and extrinsic soy effect on LDL cholesterol reduction remains the best explanation for discrepancies reported in clinical trials and meta-analyses. In the end, the large reduction in LDL cholesterol, found in earlier studies, is real and provides scientific justification to allow companies to continue using the soy heart health claim as long as they meet the amount of soy in the health claim. A further validation to the idea that there are separate intrinsic and extrinsic soy effects responsible for lowering LDL cholesterol comes from the fact that there are six studies, none of which were included in either the AHA review or the Anderson et al meta-analysis, which intentionally substituted an average of 26.6 g soy protein from soybean products for dairy products and meat without correcting for fatty acid profile and cholesterol intake. 59,60,61,62,63,64 These studies showed an average reduction in LDL cholesterol of 4.5% for the intrinsic soy effect and a 4.1% reduction for the extrinsic soy effect. Based upon a soy displacing effect at 13 g and our earlier reference to an intrinsic effect of soy at the level of 15 g in the Anderson and Bush 2011 study, NPA recommends that the soy threshold for the FDA heart health claim be lowered to 15 g in accordance with scientific findings presented here and evidence based medicine. 59 Meredith L., Liebman M., and Graves K Alterations in plasma lipid levels resulting from tofu and cheese consumption in adult women. J Am Coll Nutr 8: Dunn C. and Liebman M Plasma lipid alterations in vegetarian males resulting from the substitution of tofu for cheese. Nutr Res 6: Kurowska E.M., Jordan J., Spence J.D., Wetmore S., Piche L.A., Radzikowski M., Dandona P. and Carroll K.K Effects of substituting dietary soybean protein and oil for milk protein and fat in subjects with hypercholesterolemia. Clin Invest Med 20: Takatsuka N., Nagata C., Kurisu Y., Inaba S., Kawakami N. and Shimizu H Hypocholesterolemic effect of soymilk supplementation with usual diet in premenopausal normolipidemic Japanese women. Prev Med 31: Welty F.K., Lee K.S., Lew N.S., and Zhou J.R Effect of soy nuts on blood pressure and lipid levels in hypertensive, prehypertensive, and normotensive postmenopausal women. Arch Intern Med 167: Gardner C.D., Messina M., Kiazand A., Morris J.L. and Franke A.A Effect of two types of soy milk and dairy milk on plasma lipids in hypercholesterolemic adults: a randomized trial. J Am Coll Nutr 26:

19 Page 19 Summary of Scientific Evidence The scientific agreement of soy protein seems to be the degree to which LDL cholesterol is reduced in clinical trials because most meta-analyses have reported an effect size which favors the treatment arm with soy protein over control. The U.S. FDA s decision to allow a heart health claim on products with 25 g soy protein was initially challenged in 1996 with a manuscript by Sacks et al. 65 While the heart health claim was based upon evidence that soy protein-containing foods lowered serum LDL cholesterol, the major basis for the challenge came from later research reporting a lack of such a large cholesterol reduction in soy feeding studies 66,67,68,69 compared with earlier studies performed by Anderson et al. in 1995, 70 which demonstrated a 12.5% reduction in LDL cholesterol for serving sizes between 1.5 to 2 oz of soy protein daily (50 g/day). These later studies performed after 1996 conflicted on the degree of LDL cholesterol reduction by soy protein with meta-analyses between reporting 4-6% reductions. 71,72,73 Furthermore, the American Heart Association s Soy Advisory attributed an 65 Sacks F.M., Lichtenstein A., van Horn L., Harris W., Kris-Etherton P., and Winston M Soy protein, isoflavones, and cardiovascular health: an American Heart Association Science Advisory for professionals from the Nutrition Committee. Circulation 113: Zhan S. and Ho S.C Meta-analysis of the effects of soy protein containing isoflavones on the lipid profile. Am J Clin Nutr 81: Reynolds K., Chin A., Lees K.A., Nguyen A., Bujnowski D., and He J.A A meta-analysis of the effect of soy protein supplementation on serum lipids. Am J Cardiol 98: Harland J.I. and Haffner T.A Systematic review, meta-analysis and regression of randomized controlled trials reporting an association between an intake of circa 25 g soya protein per day and blood cholesterol. Atherosclerosis 200: Dewell A., Hollenbeck P.L., and Hollenbeck C.B Clinical review: a critical evaluation of the role of soy protein and isoflavone supplementation in the control of plasma cholesterol concentrations. J Clin Endocrinol Metab 91: Anderson J.W., Johnstone B.M., Cook-Newell M.E Meta-analysis of the effects of soy protein intake on serum lipids. N Engl J Med 333: Zhan S. and Ho S.C Meta-analysis of the effects of soy protein containing isoflavones on the lipid profile. Am J Clin Nutr 81: Reynolds K., Chin A., Lees K.A., Nguyen A., Bujnowski D., and He J.A A meta-analysis of the effect of soy protein supplementation on serum lipids. Am J Cardiol 98: Harland J.I. and Haffner T.A Systematic review, meta-analysis and regression of randomized controlled trials reporting an association between an intake of circa 25 g soya protein per day and blood cholesterol. Atherosclerosis 200:

20 Page 20 approximately 3% LDL cholesterol reduction to protein consumption of 50 g/d soy protein with no apparent dose effect. 74 So based upon the data, the challenge was not so much about whether there was any effect by soy protein at all to lower LDL cholesterol but rather the conversation centered on the degree of LDL cholesterol reduction. The intrinsic LDL cholesterol reduction (discussed earlier) is consistent with the conclusions from the AHA Soy Advisory for soy studies that are well controlled for macronutrients. However, including the extrinsic LDL lowering effect by soy greatly increases its potential as a cholesterol-lowering food and hence its role in coronary heart disease risk reduction. In fact, both intrinsic and extrinsic effects are likely to be involved in real-life scenarios when soy foods are included in suboptimal diets as suggested by NHANES III. 75 Effect of Soy Protein Compared to Other Food Components with a Health Claim The combined intrinsic and extrinsic soy LDL lowering effect is a 10.3% reduction. A heart health claim is permitted by FDA for plant sterols and stanols, consumed at 2 g/day in daily serving levels. The LDL reduction for this serving level of plant sterols and stanols is 10%, 76 which is the exact LDL cholesterol reduction identified for the combined intrinsic and extrinsic effect of soy proteins. In addition, the intrinsic and extrinsic effects of soy and the effect of plant sterols and stanol esters are much greater than other cholesterol-lowering, plant-based food components. The predicted LDL cholesterol reduction for consuming 41 g/d, the same dose for which an FDA qualified health claim is permitted 77, is smaller. Previous studies have suggested that every 10 g of almonds reduces LDL cholesterol by 1%, which would translate 74 Sacks F.M., Lichtenstein A., Van Horn L., Harris W., Kris-Etherton P., and Winston, M Soy protein, isoflavones, and cardiovascular health. An American Heart Association Science Advisory for Professionals from the Nutrition Committee. Circulation 113: National Cancer Institute. Selected intakes as ratios of energy intake. US population, Risk factor monitoring and methods branch Web site. Applied Research Program FDA Talk Paper FDA authorizes new coronary heart disease health claim for plant sterol and plant stanol esters. 5 September Washington, DC. 77 U.S. FDA FDA authorizes heart disease health claim for nuts. FDA Docket No. 02P Washington, DC.