[Federal Register Volume 63, Number 217 (Tuesday, November 10, 1998)]
[Proposed Rules]
[Pages 62977-63015]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 98-30008]
=======================================================================
-----------------------------------------------------------------------
DEPARTMENT OF HEALTH AND HUMAN SERVICES
Food and Drug Administration
21 CFR Part 101
[Docket No. 98P-0683]
Food Labeling: Health Claims; Soy Protein and Coronary Heart
Disease
AGENCY: Food and Drug Administration, HHS.
ACTION: Proposed rule.
-----------------------------------------------------------------------
SUMMARY: The Food and Drug Administration (FDA) is proposing to
authorize the use, on food labels and in food labeling, of health
claims on the association between soy protein and reduced risk of
coronary heart disease (CHD). FDA is proposing this action in response
to a petition filed by Protein Technologies International, Inc. (the
petitioner). The agency has tentatively concluded that, based on the
totality of publicly available scientific evidence, soy protein
included in a diet low in saturated fat and cholesterol may reduce the
risk of CHD.
DATES: Written comments by January 25, 1999.
ADDRESSES: Submit written comments to the Dockets Management Branch
(HFA-305), Food and Drug Administration, 5630 Fishers Lane, rm. 1061,
Rockville, MD 20852.
FOR FURTHER INFORMATION CONTACT: Susan M. Pilch, Center for Food Safety
and Applied Nutrition (HFS-465), Food and Drug Administration, 200 C
St. SW., Washington, DC 20204, 202-205-4500.
SUPPLEMENTARY INFORMATION:
I. Background
On November 8, 1990, the President signed into law the Nutrition
Labeling and Education Act of 1990 (the 1990 amendments) (Pub. L. 101-
535). This new law amended the Federal Food, Drug, and Cosmetic Act
(the act) in a number of important ways. One of the most notable
aspects of the 1990 amendments was that they provided procedures
whereby FDA is to regulate health claims on food labels and in food
labeling.
In the Federal Register of January 6, 1993 (58 FR 2478), FDA issued
a final rule that implemented the health claim provisions of the act
(hereinafter referred to as the 1993 health claims final rule). In that
final rule, FDA adopted Sec. 101.14 (21 CFR 101.14), which sets out the
rules for the authorization and use of health claims. Additionally,
Sec. 101.70 (21 CFR 101.70) establishes a process for petitioning the
agency to authorize health claims about a substance-disease
relationship (Sec. 101.70(a)) and sets out the types of information
that any such petition must include (Sec. 101.70(d)). These regulations
became effective on May 8, 1993.
In response to the 1990 amendments, FDA also conducted an extensive
review of the evidence on the 10 substance-disease relationships listed
in the 1990 amendments. As a result of its review, FDA has authorized
claims for 8 of these 10 relationships, one of which focused on the
relationship between dietary saturated fat and cholesterol and reduced
risk of CHD. CHD is the most common, most frequently reported, and most
serious form of cardiovascular disease (CVD) (58 FR 2739, January 6,
1993). Further, while the agency denied the use on food labeling of
health claims relating dietary fiber to reduced risk of CVD (58 FR
2552), it authorized a health claim relating diets low in saturated fat
and cholesterol and high in fruits, vegetables, and grain products that
contain dietary fiber (particularly soluble fiber) to a reduced risk of
CHD.
In the proposed rule entitled ``Health Claims and Label Statements;
Lipids and Cardiovascular Disease'' (56 FR 60727, November 27, 1991),
FDA set out the criteria for evaluating evidence on diet and CVD
relationships. The agency focused on those aspects of the dietary lipid
and CVD relationship for which the strongest scientific evidence and
agreement existed. FDA noted that, because of the public health
importance of CHD, identification of ``modifiable'' risk factors for
CHD had been the subject of considerable research and public policy
attention. The agency also noted that there is general agreement that
elevated blood cholesterol levels are one of the major ``modifiable''
risk factors in the development of CHD. FDA cited Federal Government
and other reviews that concluded that there is substantial
epidemiologic and clinical evidence that high blood levels of total and
low density lipoprotein (LDL) cholesterol are a cause of
atherosclerosis and represent major contributors to CHD. Further,
factors that decrease total blood cholesterol and LDL-cholesterol will
also decrease the risk of CHD. FDA concluded that it is generally
accepted that blood total and LDL-cholesterol levels are major risk
factors for CHD, and that dietary factors affecting blood cholesterol
levels affect the risk of CHD. High intakes of dietary saturated fat
and, to a lesser degree, of dietary cholesterol are consistently
associated with elevated blood cholesterol levels. FDA concluded that
the publicly available data supported an association between diets low
in saturated fat and cholesterol and reduced risk of CHD (58 FR 2739 at
2751).
Based on its review using the stated criteria, and on its
consideration of comments received in response to the proposed rule
entitled ``Health Claims; Dietary Fiber and Cardiovascular Disease''
(56 FR 60582), FDA concluded that the publicly available scientific
information supported an association between diets low in saturated fat
and cholesterol and high in fruits, vegetables, and grain products
(i.e., foods that are low in saturated fat and cholesterol and that are
good sources of dietary fiber) and reduced risk of heart disease (58 FR
2552 at 2572). In the 1993 dietary fiber and CVD final rule, in
response to a comment regarding the apparent hypocholesterolemic
properties of specific food fibers, FDA again articulated its criteria
for evaluating diet and CHD relationships (58 FR 2552 at 2567). FDA
agreed that
[[Page 62978]]
the effectiveness of naturally occurring fibers in foods in reducing
the risk of CHD may be documented for specific food products. Further,
the agency indicated that if manufacturers could document, through
appropriate studies, that dietary consumption of the soluble fiber in a
particular food has a beneficial effect on blood lipids predictive of
CHD risk, they should petition for a health claim for that particular
product. In response to two petitions that documented such evidence,
FDA has authorized health claims for soluble fiber from certain foods
and reduced risk of CHD in Sec. 101.81 (21 CFR 101.81) (62 FR 3584 at
3600, January 23, 1997, and amended at 62 FR 15343 at 15344, March 31,
1997, and 62 FR 8119, February 18, 1998).
The present rulemaking is in response to a manufacturer's health
claim petition on the relationship between soy protein and the risk of
CHD.
II. Petition for Soy Protein and Reduced Risk of CHD
A. Background
On May 4, 1998, Protein Technologies International, Inc., submitted
a health claim petition to FDA requesting that the agency authorize a
health claim on the relationship between consumption of soy protein and
the risk of CHD (Refs. 1 and 2). On August 12, 1998, the agency sent
the petitioner a letter stating that it had completed its initial
review of the petition, and that the petition would be filed in
accordance with section 403(r)(4) of the act (21 U.S.C. 343(r)(4))
(Ref. 3). In this proposed rule, the agency presents the rationale for
a health claim on this food-disease relationship as provided for under
the standard in section 403(r)(3)(B)(i) of the act and Sec. 101.14(c)
of FDA's regulations.
B. Review of Preliminary Requirements for a Health Claim
1. The Substance Is Associated With a Disease for Which the U.S.
Population Is at Risk
Several previous rules establish that CHD is a disease for which
the U.S. population is at risk, specifically claims for dietary
saturated fat and cholesterol and risk of CHD (Sec. 101.75 (21 CFR
101.75)); fruits, vegetables, and grain products and risk of CHD
(Sec. 101.77 (21 CFR 101.77)); and soluble fiber from certain foods and
risk of CHD (Sec. 101.81). FDA stated in these rules that CHD remains a
major public health problem and the number one cause of death in the
United States. Despite the decline in deaths from CHD over the past 30
years, this disease is still exacting a tremendous toll in morbidity
and mortality (Refs. 4 through 6). There are more than 500,000 deaths
each year for which CHD is an underlying cause, and another 250,000
deaths for which CHD is a contributing cause. About 20 percent of
adults (male and female; black and white) ages 20 to 74 years have
blood total cholesterol (or serum cholesterol) levels in the ``high
risk'' category (total cholesterol greater than (>) 240 milligrams (mg)
per (/) deciliter (dL) and LDL-cholesterol greater than 160 mg/dL)
(Ref. 7). Another 31 percent have ``borderline high'' cholesterol
levels (total cholesterol between 200 and 239 mg/dL and LDL-cholesterol
between 130 and 159 mg/dL) in combination with two or more risk
factors.
CHD has a significant effect on health-care costs. In 1985, total
direct costs related to CHD were estimated at $13 billion, and indirect
costs from loss of productivity due to illness, disability, and
premature deaths from this disease were an estimated $36 billion (Ref.
4). Based on these facts, FDA tentatively concludes that, as required
in Sec. 101.14(b)(1), CHD is a disease for which the U.S. population is
at risk.
2. The Substance Is a Food
The substance that is the subject of this rulemaking is soy protein
(Ref. 1). Soy protein is an edible component of the soybean, Glycine
max. Soybeans are a significant source of low-cost, high-quality
protein in the human diet.
Soy protein is used as an ingredient in other foods. It 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 upon the particular steps used during
processing, soy protein ingredients may take the form of isolated soy
protein (ISP), soy protein concentrate (SPC), or soy flour (SF). Each
of these ingredients may be further processed into texturized soy
protein or texturized vegetable protein (TVP), used in the manufacture
of meat and poultry analogs, 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 employed determine the extent of retention of such
naturally occurring constituents in the final product.
Soy protein is also consumed in the diet as a component of
traditional fermented and nonfermented soy foods such as tofu, tempeh,
and miso, in addition to whole soybeans, soynuts, soy milk, 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 their production.
Soy protein ingredients (ISP, SPC, and SF) and soy protein-
containing foods may partially replace or be used in addition to animal
or other vegetable protein sources in the human diet. Therefore, FDA
has tentatively concluded that the substance satisfies the preliminary
requirement of Sec. 101.14(b)(3)(i).
3. The Substance Is Safe and Lawful
The petitioner stated that soy protein ingredients were in common
use in food before January 1, 1958, and that they are generally
recognized as safe (GRAS) by self-determination (Ref. 1). Because the
fractionation procedures used to convert vegetable flours to vegetable
protein isolates and concentrates were commonplace prior to 1958, the
petitioner asserted that ISP and SPC can be defined as soy flour
``subject only to conventional processing as practiced prior to January
1, 1958.'' The petitioner alluded to statements that it attributed to
FDA about the GRAS status of soy protein products. (In point of fact,
however, in one document (35 FR 18530, December 5, 1970), FDA was
restating a petitioner's grounds for its petition, and in the other
document (43 FR 30472, July 14, 1978), FDA was stating a condition on
the vegetable protein products to which the proposed regulation
applied, and was not itself determining the safety or suitability of
any product (43 FR 30472 at 30474 to 30475 (comment 10).) The
petitioner also referred to unidentified statements by the U.S.
Department of Agriculture, the Association of American Feed Control
Officials, and the Codex Alimentarius that it asserted support for the
GRAS status of soy protein products (Ref. 1).
The petition also addressed some concerns that have been raised
about the potential risk of consuming soy products: Allergenicity,
exposure to trypsin inhibitors, reduced bioavailability of minerals,
and hormonal disturbances.
As is true for any protein entering the gastrointestinal tract, soy
protein has the potential to elicit an allergic reaction. Food
allergies most commonly develop in infants and young children. Although
the use of heat or hot aqueous ethanol in the processing of soybeans
destroys the immunochemical reactivity of most of the protein, a small
number of infants fed soy formula experience allergic reactions to soy
(Ref. 9). Such sensitization appears to be a manifestation of an
immature digestive tract and is rarely seen in children more
[[Page 62979]]
than 4 years old or adults. Many children outgrow food allergies (Ref.
10) and soy and seafood allergies are among those likely to be
outgrown, in contrast to allergies to milk, egg white, or peanuts.
Concerns have been raised in the past about exposure to trypsin
inhibitors contained in soybeans because these compounds had been found
to stimulate pancreatic hyperplasia and hypertrophy in animals (Ref.
11). These concerns have been allayed because heat treatment removes
most of the activity of these proteases (Ref. 12). In addition, recent
studies have questioned the applicability of the animal models, which
differ from humans in the type of diet, sensitivity of the pancreas to
trypsin inhibitors, and the anatomic sites of pancreatic cell
proliferation (Refs. 12 through 15) and have found low rates of cancer
in populations with dietary patterns that include soy foods (Ref. 16).
Soybeans contain phytic acid and dietary fiber, which have well
documented effects on reducing the bioavailability of divalent
minerals, and these components are retained in the protein fraction in
variable amounts depending upon processing. In general, the
bioavailability of minerals is lower from plant sources than animal
sources, but soy has not been found to reduce the availability of
minerals from other dietary sources consumed concurrently (Ref. 17).
Data on the possible deleterious effects of soy, and particularly its
phytate content, on mineral balance have been obtained mainly from
studies of animal models; findings in humans are less consistent and
suggest that although absorption may be impaired, overall mineral
balance is not adversely affected (Refs. 13, 18, 19, 20).
Finally, the possibility of hormonal disturbances from the weakly
estrogenic-antiestrogenic effects of soy isoflavones has been raised.
For example, infertility was found in sheep that had consumed clover
containing isoflavones (Ref. 21); however, studies of soy isoflavones
in primates showed no effects on male or female reproductive tissue or
ability (Refs. 22 through 24). Soy isoflavones have been hypothesized
as a protective factor against breast cancer in populations that
consume large amounts of soy protein (Ref. 25), and in one controlled
human trial, a 45-mg/day dose of isoflavones lead to favorable changes
in menstrual cycle length and hormone levels similar to those seen in
women treated with tamoxifen (Ref. 26).
Based on the totality of the evidence and, in particular, its
common use in food, the agency is not prepared, at this time, to take
issue with the petitioner's view that the use of soy protein is safe
and lawful as required in Sec. 101.14(b)(3)(ii). Thus, FDA tentatively
concludes that the petitioner has provided evidence that satisfies the
requirement in Sec. 101.14(b)(3)(ii) that use of soy protein at the
levels necessary to justify a claim is safe and lawful.
III. Review of Scientific Evidence
A. Basis for Evaluating the Relationship Between Soy Protein and CHD
The review examined the relationship between soy protein and CHD by
focusing on the effects of dietary intake of this substance on blood
lipid levels and on the risk of developing CHD. In the 1991 lipids-CVD
and dietary fiber-CVD health claim proposals, the agency set forth the
basis for the relationship between dietary substances and CVD (56 FR
60727 at 60728 and 56 FR 60582 at 60583). In those documents, the
agency stated that there are many risk factors that contribute to the
development of CVD, and specifically CHD, one of the most serious forms
of CVD and among the leading causes of death and disability. The agency
also stated that there is general agreement that elevated blood
cholesterol levels are one of the major ``modifiable'' risk factors in
the development of CVD and, more specifically, CHD.
The Federal Government and others who have reviewed the matter have
concluded that there is substantial epidemiologic evidence that high
blood levels of total cholesterol and LDL-cholesterol are a cause of
atherosclerosis (inadequate circulation of blood to the heart due to
narrowing of the arteries) and represent major contributors to CHD (56
FR 60727 at 60728, 56 FR 60582 at 60583, Refs. 4 through 6). Factors
that decrease total cholesterol and LDL-cholesterol will also tend to
decrease the risk of CHD. High intakes of saturated fat and, to a
lesser degree, of dietary cholesterol are associated with elevated
blood total and LDL-cholesterol levels (56 FR 60727 at 60728). Thus, it
is generally accepted that blood total cholesterol and LDL-cholesterol
levels can influence the risk of developing CHD, and, therefore, that
dietary factors affecting these blood cholesterol levels affect the
risk of CHD (Refs. 4 through 6).
When considering the effect that the diet or components of the diet
have on blood (or serum) lipids, it is also useful to consider the
effect that these factors may have on blood levels of high density
lipoprotein (HDL)-cholesterol. HDL-cholesterol appears to have a
protective effect because it is involved in the regulation of
cholesterol transport out of cells and to the liver, from which it is
ultimately excreted (Refs. 4 and 8).
For these reasons, the agency based its evaluation of the
relationship between consumption of soy protein and CHD primarily on
changes in blood total and LDL-cholesterol resulting from dietary
intervention with soy protein-containing products. A secondary
consideration was that beneficial changes in total and LDL-cholesterol
should not be accompanied by potentially adverse changes in HDL-
cholesterol. This focus is consistent with that used by the agency in
response to the 1990 amendments in deciding on the dietary saturated
fat and cholesterol and CHD health claim, Sec. 101.75 (56 FR 60727 and
58 FR 2739); the fruits, vegetables, and grain products and CHD claim,
Sec. 101.77 (56 FR 60582 and 58 FR 2552); and the soluble fiber from
certain foods and CHD claim, Sec. 101.81 (61 FR 296, 62 FR 3584, 62 FR
28234, and 63 FR 8119).
B. Review of Scientific Evidence
1. Evidence Considered in Reaching the Decision
The petitioner submitted scientific studies (Refs. 27 through 66)
evaluating the relationship between soy protein in the diet and serum
lipid levels in humans (Refs. 1 and 2). The studies submitted were
conducted between 1976 and 1998. The petition included tables that
summarized the outcome of the studies and a summary of the evidence. In
the approach taken previously in the diet and CVD proposed rules, the
agency began its review of scientific evidence in support of a health
claim by considering those studies that were published since 1988, the
date of publication of the ``Surgeon General's Report on Nutrition and
Health,'' which is the most recent and comprehensive Federal review of
the scientific evidence on dietary factors and CVD. In a brief
discussion of the role of protein in coronary heart disease, the
Surgeon General's report noted that studies of the substitution of soy
protein and other vegetable proteins for animal protein in the diets of
hyperlipidemic patients have shown a marked reduction in serum
cholesterol levels but only a small change in persons with normal
cholesterol levels (Ref. 4). Because of the brevity of this
consideration of soy protein, the agency reviewed all of the studies on
soy protein submitted by the petitioner, including those published
prior to 1988.
The petition also presented some findings from studies that
employed animal models and from in vitro
[[Page 62980]]
experiments. Human studies are weighted most heavily in the evaluation
of evidence on a diet and disease relationship; animal model and in
vitro studies can be considered as supporting evidence but cannot, in
the absence of human studies, serve as the basis for establishing that
a diet and disease relationship exists. Such studies may be useful in
providing information on the mechanism of action of soy protein's
effects on blood cholesterol levels.
2. Criteria for Selection of Human Studies
The criteria that the agency used to select the most pertinent
studies were consistent with those that the agency used to evaluate the
relationship between other substances and CHD. These criteria were that
the studies: (1) Present data and adequate descriptions of the study
design and methods; (2) be available in English; (3) include estimates
of, or enough information to estimate, soy protein intakes; (4) include
direct measurement of blood total cholesterol and other blood lipids
related to CHD; and (5) be conducted in persons who represent the
general U.S. population. In the case of (5), these persons can be
considered to be adults with blood total cholesterol levels less than
300 mg/dL. Studies of special population groups, such as adults with
very high serum cholesterol (mean greater than 300 mg/dL) and children
with hypercholesterolemia, were considered relative to the nature of
the support they provided for evidence of effect seen in studies of
subjects more representative of the general U.S. population.
In a previous rulemaking (62 FR 28234 at 28238 and 63 FR 8103 at
8107), the agency concluded that hypercholesterolemic study populations
are relevant to the general population because, based on data from the
National Health and Nutrition Examination Surveys (NHANES) III, the
prevalence of individuals with elevated blood cholesterol (i.e., 200
mg/dL or greater) is high, i.e., approximately 51 percent of adults
(Ref. 7). The proportion of adults having moderately elevated blood
cholesterol levels (i.e., between 200 and 239 mg/dL) was estimated to
be approximately 31 percent, and the proportion of adults with high
blood cholesterol levels (240 mg/dL or greater) was estimated to be
approximately 20 percent (Ref. 7). It is also estimated that 52 million
Americans 20 years of age and older would be candidates for dietary
intervention to lower blood cholesterol (Ref. 7). As the leading cause
of death in this country, CHD is a disease for which the general U.S.
population is at risk. The risk of dying from CHD is related to serum
cholesterol levels in a continuous and positive manner, increasing
slowly for levels between 150 mg/dL and 200 mg/dL and more rapidly when
the cholesterol level exceeds 200 mg/dL (Ref. 67). The public health
policy elucidated by the National Cholesterol Education Program (NCEP),
National Heart, Lung, and Blood Institute, is to extend the benefits of
cholesterol lowering to the population as a whole by promoting adoption
of eating patterns that can help lower the blood cholesterol levels of
most Americans (Ref. 67). A dietary intervention that lowers blood
cholesterol levels mainly or only in persons with high levels would,
like an intervention that lowers cholesterol levels across the entire
population range, cause a shift in the population distribution of blood
cholesterol levels resulting in a decrease in the mean value for the
blood cholesterol level in the general population (Ref. 67). The
anticipated effect of such a shift would be to reduce the morbidity
from CHD and to produce a continued or accelerated decline in the CHD
mortality rate in the United States. Accordingly, in this proposal, the
agency has reviewed and considered the evidence of effects of soy
protein on serum lipids in hypercholesterolemic subjects.
In selecting human studies for review, the agency excluded studies
that were published in abstract form because they lacked sufficient
detail on study design and methodologies, and because they could not
provide the primary data.
3. Criteria for Evaluating the Relationship Between Soy Protein and CHD
Well reasoned approaches for evaluating studies supporting diet/
disease relationships are summarized in the comprehensive report ``Diet
and Health'' issued by the National Academy of Sciences (Ref. 68) and
``The Guide to Clinical Preventive Services'' issued by the U.S.
Preventive Services Task Force (Ref. 69). The criteria articulated in
these documents provided a starting point for FDA's review of
individual studies on the relationship between dietary factors and CHD
in previous rulemakings: In the 1991 proposed rule on lipids and CVD
(56 FR 60727), in the 1991 proposed rule on dietary fiber and CVD (56
FR 60582), in the January 1996 proposed rule on whole oats and CHD (61
FR 296), and in the May 22, 1997, proposed rule on soluble fiber from
psyllium and CHD (62 FR 28234).
The criteria that the agency used in evaluating the studies for
this rulemaking include: (1) Reliability and accuracy of the methods
used in nutrient intake analysis, including measurements of soy protein
intake; (2) estimates of intake of saturated fat and cholesterol; (3)
available information on the soy protein test products and control
foods; (4) measurement of study endpoints (i.e., measurement of blood
lipid levels); and (5) general study design characteristics.
The general study design characteristics for which the agency
looked included randomization of subjects, appropriateness of controls,
selection criteria for subjects, attrition rates (including reasons for
attrition), potential for misclassification of individuals with regard
to dietary intakes, presence of recall bias and interviewer bias,
recognition and control of confounding factors (for example, monitoring
body weight and control of weight loss), appropriateness of statistical
tests and comparisons, and statistical power of the studies. The agency
considered whether the intervention studies that it evaluated had been
of long enough duration, greater than or equal to 3 weeks duration, to
ensure reasonable stabilization of blood lipids.
C. Review of Human Studies
FDA conducted a comprehensive review of 41 of 43 human intervention
studies submitted in the petition and reported in 38 references by the
petitioner (Refs. 27 through 64). The two studies FDA excluded from
consideration at the outset (Refs. 32 and 52) were of infants. Of the
studies reviewed, 27 met the aforementioned criteria for selection
(Refs. 27, 28, 29, 30 (1 trial), 31, 33, 34, 35, 36, 37, 40 (2 trials),
42 and 45 (1 trial), 43, 44, 46, 49, 51, 53, 54, 55, 56, 58, 59, 60,
63, and 64). Of these, the agency gave particular weight to 14 trials
(Refs. 27, 28, 30 (1 trial), 31, 36, 37 (1 trial), 40 (2 trials), 44,
49, 51, 54, 58, and 59) that included subjects representative of the
general U.S. population and that were well controlled, reported intakes
of saturated fat and cholesterol, and avoided problems associated with
small sample size, lack of a placebo, and other design problems. These
studies are summarized in Table 1 at the end of this document and
discussed in section III.C.1 of this document. Three additional similar
trials that were included in the review but accorded less weight
because of issues concerning the populations studied and diets fed
(Refs. 29, 43, and 53) are also summarized in Table 1 of this document
and discussed in section III.C.1 of this document. Seven trials in
adults (Refs. 33, 35, 46,
[[Page 62981]]
55, 56, 60, and 64) and three trials in children (Refs. 34, 42 and 45
(1 trial), and 63) with type II or familial hypercholesterolemia are
summarized in Table 2 at the end of this document and discussed in
section III.C.2 of this document. The fourteen remaining intervention
trials (Refs. 30 (1 trial), 37 (1 trail), 38, 39 (2 trials), 41, 47,
48, 50 (2 trials), 57, 61, and 62 (2 trials)) failed to meet the
inclusion criteria because of small sample size, inadequate period of
intervention, inadequate characterization of the soy protein tested,
inadequate information on dietary intake, or lack of data on outcome
variables. The results of one epidemiological study (Ref. 65) and a
meta-analysis (Ref. 66) that included a number of the soy protein
studies submitted in the petition are discussed in sections III.C.3 and
III.C.4, respectively, of this document.
1. Studies of Adult Subjects Representative of the General U.S.
Population (Serum Cholesterol <300 mg/dL)
The agency began its consideration of the data with the 14 well
controlled and representative studies identified previously (Refs. 27,
28, 30 (1 trial), 31, 36, 37 (1 trial), 40 (2 trials), 44, 49, 51, 54,
58, and 59). Several of these studies examined the interaction of
protein and other components of soy protein sources hypothesized to
have an impact on lipid-lowering effects (i.e., isoflavones, dietary
fiber, and soy lipids) (Refs. 31, 28, 27, 51, and 44). Findings with
respect to soy protein are described in this section, while findings
regarding the specific influence of soy isoflavones (Refs. 31 and 28)
are discussed in more detail in section III.C.5 of this document.
In hypercholesterolemic subjects, Crouse et al. (Ref. 31,
documented in Ref. 1 with corrections noted in Ref. 2) found that 25
grams (g) of soy protein from ISP containing 2.5 mg total aglycone
isoflavones/g protein lowered total (p<0.05) and LDL-cholesterol levels
(p<0.05) by 4 and 6 percent, respectively, while HDL-cholesterol was
not altered. Furthermore, in subjects with LDL-cholesterol levels in
the top half of the study population, serum total and LDL-cholesterol
were reduced by 9 percent (p<0.03) and 12 percent (p<0.03),
respectively, by the ISP with 2.5 mg total aglycone isoflavones/g
protein, and by 8 percent (p<0.03) and 9 percent (p<0.03),
respectively, by the ISP with 1.6 mg total aglycone isoflavones/g
protein. HDL-cholesterol concentrations were unchanged. These results
indicate that soy protein, in a diet low in saturated fat and
cholesterol, can exert hypocholesterolemic effects but suggest these
effects may be modulated by the presence of isoflavones.
In hypercholesterolemic, postmenopausal women, Baum et al. (Ref.
28) also investigated the impact of soy protein as ISP containing
different levels of isoflavones. Adjusted mean differences in the
change from baseline for total serum cholesterol level did not differ
in the two soy groups and the control group. However, there was a
statistically significant reduction of 8 to 9 percent in non-HDL (LDL
plus very low density lipoprotein (VLDL)) cholesterol in both of the
ISP treatment groups (p<0.05) compared to the control group. HDL-
cholesterol was also significantly increased (p<0.05) in both soy
groups compared to the control. The level of isoflavones did not affect
any of the blood lipid levels measured. This study also indicates the
ability of soy protein provided in a diet low in saturated fat and
cholesterol to reduce LDL-cholesterol.
Two studies that examined the effect of soy protein in
hypercholesterolemic adults consuming low fat diets also evaluated
whether soy cotyledon fiber had additional lipid-lowering effects.
Bakhit et al. (Ref. 27) used 25 g protein and 20 g dietary fiber as
treatment levels while Potter et al. (Ref. 51) used 50 g protein and 20
g dietary fiber. Soy protein was provided as ISP (Refs. 27 and 51) and
SF (Ref. 51) incorporated into baked products.
Bakhit et al. (Ref. 27) studied subjects who had initially been
screened for eligibility based on plasma total cholesterol
concentrations greater than 220 mg/dL before starting the study. During
the baseline dietary period, plasma total cholesterol decreased to
levels below 220 mg/dL in 10 of the subjects; these subjects did not
have any further decrease in total or LDL-cholesterol with any of the
experimental diets. The subjects whose cholesterol remained greater
than the 220 mg/dL intent-to-treat level did show a statistically
significant decrease from post-baseline dietary levels for total
cholesterol, but not for LDL-cholesterol, after consuming ISP. In the
subset analysis, Bakhit et al. (Ref. 27) found a statistically
significant decrease in total cholesterol of 7 percent (p<0.05) from
post-stabilization levels with ingestion of ISP. Addition of soy
cotyledon fiber to the ISP diet resulted in a statistically significant
decrease (p<0.05) of 8 percent in total cholesterol. Ingestion of the
casein plus cellulose control diet produced a nonsignificant decrease
(p>0.05) in total cholesterol of 3 percent. Differences in LDL- and
HDL-cholesterol from baseline or control after the two soy diets were
not statistically significant. In the subset analysis, the additional
effect of soy fiber on blood cholesterol levels was not significant
when evaluated by analysis of covariance (p=0.04 for protein effects;
p=0.07 for fiber effects). This study supports a conclusion that the
protein and not the fiber component of the soybean is largely
responsible for effects on blood lipids.
Potter et al. (Ref. 51) reported a statistically significant
(p<0.05) decrease in plasma total cholesterol from baseline of 8
percent with ingestion of diets containing ISP whether soy cotyledon
fiber or cellulose was also consumed. The 8-percent decrease observed
in LDL-cholesterol from baseline was statistically significant only
when the ISP diet also contained soy cotyledon fiber (p<0.05). Total
and LDL-cholesterol were also significantly (p<0.01) lower with the ISP
diets compared to the nonfat dry milk-cellulose control diet. No
statistically significant changes in HDL-cholesterol were observed with
any of the soy protein diets. Changes from baseline were not
statistically significant for any of the blood lipids when the diet
providing soy protein as SF was consumed. However, the difference in
total cholesterol observed after ingestion of SF was 19 mg/dL lower
than that on the control diet of nonfat dry milk and cellulose
(p<0.01). These findings suggest that the principal dietary component
responsible for the lipid-lowering observed in this study is the soy
protein fraction, and that soy fiber may have an incremental effect.
Kurowska et al. (Ref. 44) tested the effects of soy protein and soy
oil in hypercholesterolemic subjects by adding combinations of ``milk''
and desserts to provide a total of 31 g protein from either cow's milk
or soy milk and 16 g fat from either cow's milk, soybean oil, or whole
soybean soy milk. The three dietary treatments were cow's milk (2-
percent fat), skim cow's milk (0-percent fat) plus soy oil (16 g), or
soybean milk. No statistically significant changes from baseline in
total cholesterol were observed in response to any of the dietary
treatments. The 4-percent decline in LDL-cholesterol observed with the
soybean milk diet was not statistically significant. HDL-cholesterol
was increased 7 percent from baseline (p=0.04) with the whole soybean
milk treatment. In the subjects with the highest initial LDL-
cholesterol level and LDL/HDL-cholesterol ratio, LDL-cholesterol was
reduced by 11 percent by the soybean milk diet.
[[Page 62982]]
Five earlier studies included in Table 1 reported on effects of soy
protein in hypercholesterolemic subjects.
In hypercholesterolemic subjects, Goldberg et al. (Ref. 37)
examined the effects of ISP (99 g of soy protein) incorporated as a
meat analog or formulated in beverage compared to a control animal
protein diet consisting of analogous meat products and nonfat dry milk.
Both diets resulted in statistically significant reductions in serum
total and LDL-cholesterol levels. With the soy protein diet, total
cholesterol was decreased by 15 percent (p<0.001) and LDL-cholesterol
was decreased by 17 percent (p<0.001) from baseline values. Total
cholesterol was 8 mg/dL lower (p<0.005), and LDL-cholesterol was 10 mg/
dL lower (p<0.05), at the end of the dietary period when soy protein
was ingested as compared to the animal protein diet. Both the change in
HDL-cholesterol from the baseline and the difference in HDL-cholesterol
between the soy and control diets were small and not statistically
significant.
Mercer et al. (Ref. 49) tested the effects of approximately 17 g of
soy protein from ISP as a replacement for 2-percent fat cow's milk in
subjects with mild to moderate hypercholesterolemia. Total cholesterol
levels were not significantly different (p>0.05) on the two diets.
However, among the subjects whose baseline total cholesterol was above
the 90th percentile, the soy protein diet resulted in a decrease from
baseline in mean total cholesterol of 4 percent and a level 9 percent
lower (16 mg/dL; p<0.05) than the level at the end of the cow's milk
period. There were no statistically significant differences in LDL-
cholesterol and HDL-cholesterol between ISP and cow's milk diets either
for all subjects or for the subset of subjects with the highest initial
total cholesterol levels.
Holmes et al. (Ref. 40) conducted two trials with
hypercholesterolemic subjects testing SF as a texturized vegetable
protein product formulated with egg yolk, beef tallow, and cottonseed
oil to create an analog for lean ground beef. An average of 27 g of soy
protein was consumed in the partially substituted diet in the first
trial and 62 g was consumed in the completely substituted diet in the
second trial. In trial 1, statistically significant changes in total
cholesterol (p<0.02) and LDL-cholesterol (p<0.05) occurred during the
initial stabilization period when the control diet was consumed; no
further changes occurred after the second period during which the
partially substituted soy diet was consumed. In trial 2, both diets
significantly lowered mean total cholesterol during the first dietary
sequence (p<0.05), the animal protein diet by 18 percent and the soy
diet by 19 percent. Crossing over the diets had no further effect. LDL-
cholesterol levels were not reduced by either diet. HDL-cholesterol
levels were not significantly affected by diet in either trial. The two
trials were unique in the source of soy protein and in including
subjects with type IV hyperlipidemia.
Shorey et al. (Ref. 54) examined the effects of 57 g of soy protein
(mean intake) consumed as ISP incorporated both into meat analogs and a
soy-based beverage in hypercholesterolemic young men. A statistically
significant (p=0.027) decrease from baseline total cholesterol of 7
percent was noted in the group consuming the soy protein diet; however,
these values were 6 mg/dL higher than change from baseline values
obtained from the control group. HDL-cholesterol also significantly
(p=0.001) decreased from baseline values by 15 percent. LDL-cholesterol
was not measured in this study. Although the two diets were well
matched for saturated fat and cholesterol, interpretation of these
findings is complicated by the fact that body weight was significantly
(p<0.004) decreased in both groups of subjects. Subjects who showed a
significant hypocholesterolemic response on either diet were those who
substantially reduced their customary protein and fat intakes on the
experimental diets. In contrast to other studies, subjects in this
study with lower baseline values experienced more pronounced reductions
in total cholesterol level.
Four additional well-controlled studies included in Table 1 of this
document examined the effects of soy protein in normocholesterolemic
subjects.
The study of Carroll et al. (Ref. 30) compared ISP (44 g soy
protein estimated) incorporated into foods and a soy-based beverage to
a mixed protein/animal-based diet in healthy young women. Plasma total
cholesterol was significantly (p<0.05) lower, by 10 mg/dL, when the soy
protein diet was consumed as compared with the mixed protein diet.
Neither LDL-cholesterol nor HDL-cholesterol was measured.
Giovannetti et al. (Ref. 36) examined the effects of ISP (66 to 80
g of soy protein depending on energy intake) incorporated as meat and
dairy analogs in healthy young adult women in both high- and low-fat
diets. On the high-fat diet, serum total cholesterol was 4 mg/dL lower,
LDL-cholesterol was 6 mg/dL lower, and HDL-cholesterol was 3 mg/dL
lower after ingestion of the soy protein than after ingestion of the
mixed protein control. None of the changes in blood lipids reached
statistical significance. On the low-fat diet, serum total cholesterol
was 1 mg/dL higher, LDL-cholesterol was 5 mg/dL lower, and HDL-
cholesterol was 2 mg/dL higher after soy protein than after the mixed
protein control; these differences were not statistically significant.
The magnitude of reduction in serum total cholesterol with soy protein
was similar on the high-fat and low-fat diets, 10 percent and 9
percent, respectively. Substitution of soy protein caused reductions in
LDL-cholesterol levels during the high-fat diet in 11 of 12 subjects
and during the low-fat diet in 9 of 12 subjects.
Van Raaij et al. (Ref. 58) tested the effects of ISP in young
normocholesterolemic men and women consuming three diets that differed
in protein composition with 65 percent of the total protein replaced by
either soy protein (54 g), or casein, or an approximately 2:1 mixture
of casein (36 g):soy (17 g). In the group consuming the soy protein
diet, total serum cholesterol and LDL-cholesterol were decreased (-2
percent and -8 percent, respectively) and HDL-cholesterol increased
(+10 percent) compared to values at the end of the lead-in period. The
changes in both LDL-cholesterol and HDL-cholesterol were statistically
significant (p<0.05). In addition, decreases in LDL-cholesterol were
significantly (p<0.05) greater with the soy protein diet compared to
changes with the casein diet. Although weight loss did occur among
subjects consuming both the soy protein diet (n=9) and the casein diet
(n=6), when data from the subset without a weight loss of more than 2
kilograms (kg) were analyzed separately, the same effects of soy
protein ingestion on blood lipid-lowering were observed. The lipid
changes in the group that remained on the 2:1 casein:soy diet were not
statistically significantly different from the casein group, nor were
changes from the end of the stabilization period significant in this
group.
In a trial with both normocholesterolemic and hypercholesterolemic
subjects, Van Raaij et al. (Ref. 59) tested both ISP and SPC (each
providing an average of 55 g of soy protein) compared to a casein
control. Serum total cholesterol was decreased from baseline by 4
percent and LDL-cholesterol was decreased by 3 percent on the ISP diet.
These changes were significantly different from those on the SPC diet
(p<0.05) but not significantly different from those on the casein diet.
HDL-cholesterol showed a slight but statistically significant
[[Page 62983]]
increase of 2 percent from baseline on the ISP diet, a change that was
also significantly different from that on the casein diet. When SPC was
used as the protein source, total cholesterol was not altered, LDL-
cholesterol was increased by 6 percent, and HDL-cholesterol decreased
by 3 percent compared to baseline. None of these changes in blood
lipids from baseline or differences between the casein and SPC diets
was statistically significant. Interpretation of this study is
complicated by differential weight loss on the experimental diets
(weight loss was greatest in the casein group) and differential fiber
intake.
Three additional studies (Refs. 29, 43, and 53), in which
interpretation is complicated by design issues such as choice of
subjects, concerns about weight loss, or uncertainties about other
components in diets, are also summarized in Table 1 of this document
and discussed as follows.
Bosello et al. (Ref. 29) and Jenkins et al. (Ref. 43) both studied
the hypocholesterolemic effects of soy protein versus casein in the
context of hypocaloric diets fed to obese persons to achieve
significant weight reduction. In Bosello et al. (Ref. 29), obese
subjects (>150 percent of ideal body weight) received 375 kilocalorie
(kcal)/day initially, followed by an 800 kcal/day diet. During both
phases, the 375 kcal portion was provided by commercial textured
protein products that delivered either 27 g protein from casein or 27 g
protein from soy protein (type of soy protein not given). During the
second phase, the 375 kcal/day was ``integrated'' with an extra 425
kcal/day from conventional foods. Mean weight losses for the soy and
casein groups were 17 and 16 kg, respectively. Total cholesterol and
LDL-cholesterol in the soy group were both 16 percent lower compared to
baseline (p<0.01). Compared to the casein group, total cholesterol was
20 mg/dL lower (p<0.01 ) and LDL-cholesterol was 16 mg/dL lower
(p<0.01). HDL-cholesterol was decreased in both groups at the end of
the study; however, only in the casein group was the difference
statistically significant (p<0.01). Additionally, the decrease in HDL-
cholesterol in the casein group was significantly (p<0.01) greater than
that observed in the soy protein group.
Jenkins et al. (Ref. 43) examined the effects of soy protein
ingestion on serum cholesterol in obese women who were also consuming a
hypocaloric diet for weight reduction. The three treatments were: A
control, hypocaloric diet of 1,000 total kcal consumed as conventional
foods; the same diet with two meals per day replaced by a soy protein
(18.4 g provided as ISP) liquid formula preparation; or the same diet
with two meals per day replaced by a milk protein (17.6 g as milk
protein isolate and nonfat dry milk) liquid formula. An average 2.5 kg
weight loss per month occurred during the study (p<0.05) across diet
treatments. Statistically significant decreases from baseline in total
cholesterol of 10 percent (p<0.05) and in LDL-cholesterol of 17 percent
(p<0.05) occurred only during the period when the soy protein formula
was ingested. Changes in HDL-cholesterol were not statistically
significant. These effects of soy protein were independent of the order
the soy diet was consumed relative to the conventional hypocaloric
diet. The levels of total and LDL-cholesterol achieved with ingestion
of soy protein were, respectively, 10 mg/dL and 8 mg/dL lower with the
soy protein diet as compared with the casein diet. Neither the
conventional hypocaloric diet nor the casein formula hypocaloric diet
resulted in statistically significant decreases in total or LDL-
cholesterol despite weight loss. Calculations of the expected decline
in serum total cholesterol based on changes in weight, dietary
cholesterol, and saturated and polyunsaturated fat accurately predicted
the observed changes in both the hypocaloric diet and milk formula
groups, but significantly underestimated the decrease observed in the
soy formula group.
Sacks et al. (Ref. 53) studied the effects of 27 g of protein
consumed daily as ISP or casein incorporated into muffins and oatmeal
in adults who were strict vegetarians. Not unexpectedly, given the very
low baseline lipid concentrations and very low dietary fat and
cholesterol intake, no statistically significant changes or differences
in total cholesterol, LDL-cholesterol or HDL-cholesterol were observed
from consumption of either soy protein or casein.
a. Summary--Hypercholesterolemic subjects consuming diets low in
saturated fat and cholesterol. In five (Refs. 31, 28, 27, 51, and 44)
of seven (Refs. 31, 28, 27, 51, 44, and 40 (2 trials)) well-controlled
studies of hypercholesterolemic subjects consuming low saturated fat
and low cholesterol diets, soy protein intake was associated with
significant decreases in total and/or LDL-cholesterol levels. Crouse et
al. (Ref. 31, documented in Ref. 1 with corrections noted in Ref. 2)
found that soy protein from ISP containing 2.5 mg total aglycone
isoflavones/g protein statistically significantly lowered total
(p<0.05) and LDL-cholesterol levels (p<0.05), by 4 and 6 percent,
respectively, while HDL-cholesterol was not altered. In a subset of
subjects with LDL-cholesterol levels in the top half of the study
population, serum total and LDL-cholesterol were reduced by 9 percent
(p<0.03) and 12 percent (p<0.03), respectively, by the ISP with 2.5 mg
total aglycone isoflavones/g protein, and by 8 percent (p<0.03) and 9
percent (p<0.03), respectively, by the ISP with 1.6 mg total aglycone
isoflavones/g protein. Baum et al. (Ref. 28) found that the adjusted
mean difference in total serum cholesterol level was not significantly
(p>0.05) different in the two groups consuming soy as ISP and the
control group. However, there was a statistically significant reduction
of 8 to 9 percent in non-HDL (LDL plus VLDL) cholesterol in both of the
ISP treatment groups (p=0.04) compared to the control group.
Bakhit et al. (Ref. 27) found, in a subset of subjects whose
cholesterol remained greater than the 220 mg/dL intent-to-treat level
after run-in with the baseline diet, a statistically significant
decrease in total cholesterol of 7 percent (p<0.05) from post-
stabilization levels with ingestion of ISP; addition of soy cotyledon
fiber to the ISP diet resulted in a significant decrease (p<0.05) of 8
percent in total cholesterol. Levels of LDL-cholesterol were not
statistically significantly affected by either soy diet. Potter et al.
(Ref. 51) reported a statistically significant decrease (p<0.05) from
baseline in total plasma cholesterol of 8 percent with ingestion of
diets containing ISP whether soy cotyledon fiber or cellulose was also
consumed. The 8-percent decrease in LDL-cholesterol from baseline was
statistically significant only when the ISP diet also contained soy
cotyledon fiber (p<0.05). Total and LDL-cholesterol were also
significantly lower (p<0.01) with the ISP diets compared to the nonfat
dry milk-cellulose diet. Changes from baseline were not statistically
significant for any of the blood lipids when the diet providing soy
protein as SF was consumed. However, the difference in total
cholesterol observed after ingestion of SF was 19 mg/dL lower than that
on the control diet of nonfat dry milk and cellulose (p<0.01).
With diets providing either cow's milk (2-percent fat), or skim
cow's milk (0-percent fat) plus soy oil (16 g), or soybean milk,
Kurowska et al. (Ref. 44) found no statistically significant changes
from baseline in total cholesterol and LDL-cholesterol in response to
any of the dietary treatments. In the subjects with the highest initial
LDL-cholesterol levels and LDL/HDL-cholesterol ratios, LDL-
[[Page 62984]]
cholesterol was reduced by 11 percent by the soybean milk diet. Holmes
et al. (Ref. 40) conducted two trials testing SF as a texturized
vegetable protein product, with averages of 27 and 62 g of soy protein
consumed, respectively, in the first and the second trial. In trial 1,
statistically significant changes in total and LDL-cholesterol occurred
during the stabilization period when the control diet was consumed; no
further changes occurred after the second dietary period during which
the partially substituted soy diet was consumed. In trial 2, both diets
resulted in a statistically significant lowering of total cholesterol
during the first dietary sequence, the animal protein diet by 18
percent and the soy diet by 19 percent. Crossing over the diets had no
further effect. LDL-cholesterol levels were not reduced by either diet.
These studies were unique in the source of soy protein used and in
including subjects with type IV hyperlipidemia.
Levels of HDL-cholesterol were also measured in each of these seven
studies and were found either to be unchanged (Refs. 31, 27, 51, and 40
(2 trials)) or to show a slight but statistically significant increase
(Refs. 28 and 44) in response to consumption of diets containing soy
protein.
Levels of soy protein as ISP found to be effective in lowering
total and LDL-cholesterol levels ranged in these studies from 25 to 50
g (Refs. 31, 28, 27, and 51). As whole soybean milk, 31 g of soy
protein lowered LDL-cholesterol only in the subset of subjects with the
highest initial LDL-cholesterol levels and LDL/HDL-cholesterol levels
(Ref. 44). Diets providing 50 g of soy protein as SF did not cause
significant changes from baseline for any of the blood lipids, but the
decrease in total cholesterol observed after ingestion of SF was
significantly greater than that on the control diet of nonfat dry milk
and cellulose (Ref. 51). Diets providing 27 g of soy protein as SF in a
textured product had no significant effects on blood lipid levels
compared to a control diet, and a higher level (62 g) significantly
lowered total cholesterol only in the experimental group fed the soy
protein diet first (Ref. 40).
b. Summary--Hypercholesterolemic subjects consuming ``usual''
diets. Three studies reported on effects of soy protein in
hypercholesterolemic subjects consuming ``usual'' diets that were
generally high in total fat, saturated fat, and cholesterol (Refs. 37,
49, and 54). Goldberg et al. (Ref. 37) found, on the soy protein diet
(with 99 g of soy protein as ISP), statistically significant decreases
from baseline of 15 percent in total cholesterol and 17 percent in LDL-
cholesterol. Total cholesterol was 8 mg/dL lower (p<0.005), and LDL-
cholesterol was 10 mg/dL lower (p<0.05), at the end of the dietary
period when soy protein was ingested as compared to the animal protein
diet. Mercer et al. (Ref. 49) found that a diet with approximately 17 g
of soy protein from ISP did not produce changes in serum cholesterol
that were significantly different from those of a cow's milk control
diet. Among subjects whose baseline total cholesterol was above the
90th percentile, Mercer et al. (Ref. 49) found that the soy protein
diet resulted in a decrease from baseline in mean total cholesterol of
4 percent and a level 9 percent lower (16 mg/dL; p<0.05) than the level
at the end of the cow's milk control period. LDL-cholesterol did not
differ significantly between ISP and cow's milk diets for all subjects
or for the subset of subjects with the highest initial total
cholesterol levels.
Shorey et al. (Ref. 54) found diets with 57 g of soy protein as ISP
was associated with a statistically significant decrease from baseline
in total cholesterol of 7 percent (p=0.027); however, these values were
6 mg/dL higher than change from baseline values obtained from the
control group. LDL-cholesterol was not measured in this study. Although
the two diets were well matched for saturated fat and cholesterol,
interpretation of these findings is complicated by the fact that body
weight was significantly decreased in both groups of subjects
(p<0.004). Subjects who showed a significant hypocholesterolemic
response on either diet were those who substantially reduced their
customary protein and fat intakes on the experimental diets. In
contrast to other studies, subjects in this study with lower baseline
values experienced more pronounced reductions in total cholesterol
level.
HDL-cholesterol was also measured in these three studies. Changes
were small and not statistically significant in two studies (Refs. 37
and 49), but HDL-cholesterol was significantly decreased from baseline
values by 15 percent in one study (Ref. 54). (This latter study had a
number of anomalous results.)
Each of these three studies fed soy protein in experimental diets
as ISP (Refs. 37, 49, and 54). With a diet containing a very high level
(99 g) of soy protein from this source (Ref. 37), statistically
significant differences in both total and LDL-cholesterol were
reported. Results were less consistent with a relatively low level of
soy protein (17 g) (Ref. 49). An intermediate level of soy protein (57
g) was found to be ineffective in lowering total cholesterol in the
study of Shorey et al. (Ref. 54).
c. Summary--Normocholesterolemic subjects. Five studies examined
the effects of soy protein in normocholesterolemic subjects (Refs. 30,
36, 58, 59, and 53). The study of Carroll et al. (Ref. 30) found plasma
total cholesterol was significantly lower (-10 mg/dL) when a soy
protein diet (low in saturated fat and cholesterol and providing an
estimated 44 g soy protein as ISP) was consumed as compared with a
mixed protein control diet (p<0.05). LDL-cholesterol was not measured.
Giovannetti et al. (Ref. 36) examined the effects of soy protein as ISP
(66 to 80 g of soy protein depending on energy intake) in both high-
and low-fat diets. Changes in total and LDL-cholesterol with the soy
protein diets were not statistically significantly different from
changes with the corresponding control diets, regardless of fat
content. The magnitude of reduction in serum total cholesterol with soy
protein was similar on the high-fat and low-fat diets, 10 percent and 9
percent, respectively. Substitution of soy protein caused reductions in
LDL-cholesterol levels during the high-fat diet in 11 of 12 subjects
and during the low-fat diet in 9 of 12 subjects.
Van Raaij et al. (Ref. 58) tested the effects of ISP using three
diets high in total fat, saturated fat, and cholesterol that differed
in protein composition with 65 percent of the total protein comprising
either soy protein (54 g), or casein, or an approximately 2:1 mixture
of casein (36 g):soy (17 g). In the group consuming the soy protein
diet, the decrease in total serum cholesterol (-2 percent) was not
statistically significant, but the decrease in LDL-cholesterol (-8
percent) was statistically significant (p<0.05). In addition, decreases
in LDL-cholesterol were significantly greater with the soy protein diet
compared to changes with the casein diet (p<0.05).
In a trial with both normocholesterolemic and moderately
hypercholesterolemic subjects, Van Raaij et al. (Ref. 59) tested both
ISP and SPC (each providing an average of 55 g of soy protein) compared
to a casein control in diets high in total fat, saturated fat, and
cholesterol. Serum total cholesterol was decreased from baseline by 4
percent and LDL-cholesterol was decreased by 3 percent on the ISP diet.
These changes were statistically significantly different from those on
the SPC diet (p<0.05) but not significantly different from those on the
casein diet. When SPC was used as the protein source, total cholesterol
was not altered and LDL-cholesterol was increased by 6 percent compared
to
[[Page 62985]]
baseline. None of these changes in blood lipids from baseline or
differences between the casein and SPC diets was statistically
significant. Interpretation of this study is complicated by
differential weight loss on the experimental diets (weight loss was
greatest in the casein group) and differential fiber intake.
Sacks et al. (Ref. 53) studied the effects of 27 g of protein
consumed daily as ISP or casein incorporated into muffins and oatmeal,
in diets very low in saturated fat and cholesterol in adults who were
strict vegetarians. Not unexpectedly, given the very low baseline lipid
concentrations and very low dietary fat and cholesterol intake, no
statistically significant changes or differences in total cholesterol
or LDL-cholesterol or HDL-cholesterol were observed from consumption of
either soy protein or casein.
HDL-cholesterol was measured in four of these studies, with
statistically significant increases associated with soy protein intake
found in two (Refs. 58 and 59) and no statistically significant changes
in two (Refs. 36 and 53).
Effects of soy protein on total and LDL-cholesterol were less
consistent in normocholesterolemic subjects than in moderately
hypercholesterolemic subjects. As ISP, 44 g of soy protein was
effective in statistically significantly lowering total cholesterol in
one study (Ref. 30), and 54 g statistically significantly lowered LDL-
cholesterol in one study (Ref. 58). With very low initial blood lipid
levels, the impact of dietary changes appears to be lessened.
d. Summary--Subjects consuming hypocaloric diets. Bosello et al.
(Ref. 29) and Jenkins et al. (Ref. 43) both studied the
hypocholesterolemic effects of soy protein versus casein in the context
of hypocaloric diets fed to obese persons to achieve significant weight
reduction. In Bosello et al. (Ref. 29), total cholesterol and LDL-
cholesterol in the soy group (which consumed 27 g of soy protein) were
both 16 percent lower compared to baseline (p<0.01). Compared to the
casein control group, total cholesterol was 20 mg/dL lower (p<0.01 )
and LDL-cholesterol was 16 mg/dL lower (p<0.01) in the soy protein
group. Jenkins et al. (Ref. 43) found that statistically significant
decreases from baseline in total cholesterol of 10 percent (p<0.05) and
in LDL-cholesterol of 17 percent (p<0.05) occurred only during the
period when the soy protein formula (which provided 17 g of soy
protein) was ingested. The levels of total and LDL-cholesterol achieved
with ingestion of soy protein were, respectively, 10 mg/dL and 8 mg/dL
lower with the soy protein diet compared with casein diet. Neither the
conventional hypocaloric diet nor the casein formula hypocaloric diet
resulted in statistically significant decreases in total or LDL-
cholesterol despite weight loss.
HDL-cholesterol was decreased in both groups at the end of the
first study (Ref. 29); however, only the casein group's values were
significantly (p<0.01) different from baseline. Additionally, the
decrease in HDL-cholesterol in the casein group was
significantly(p<0.01) greater than that observed in the soy protein
group. In the second study (Ref. 43), HDL-cholesterol levels were not
significantly affected by dietary treatment.
These two studies (Refs. 29 and 43) demonstrated decreases in both
total and LDL-cholesterol levels during hypocaloric diets that provided
relatively low amounts (27 and 17 g, respectively) of soy protein.
2. Studies of Subjects with Type II and Familial Hypercholesterolemia
(Mean Total Cholesterol Level>300 mg/dL)
Ten studies (Refs. 33, 35, 46, 55, 56, 60, 64, 34, 42 and 45 (1
trial), and 63) of subjects with severe (type II or familial)
hypercholesterolemia (mean total cholesterol level>300 mg/dL) are
summarized in Table 2 of this document and discussed in section III.
C.2 of this document. Seven report results in adults (Refs. 33, 35, 46,
55, 56, 60, and 64) and three in children (Refs. 34, 42 and 45 (1
trial), and 63).
a. Studies in adults. Sirtori et al. (Ref. 55) reported a decrease
of 21 percent in both total (p<0.001) and LDL-cholesterol (p<0.01) with
soy protein consumption in adults with type II hyperlipoproteinemia.
Total intake of soy protein, as a textured protein isolate, was not
given but was approximately 13 percent of kcal or 60 g. The order in
which the soy protein diet was consumed did not affect the results and
the changes in total plasma cholesterol level far exceeded those
expected based on the small differences in ratio of polyunsaturated to
saturated fat and cholesterol content of the diets. When the control
diet was fed first, statistically significant changes in total and LDL-
cholesterol were not observed; when it was fed second, total
cholesterol increased statistically significantly. These investigators
also reported that addition of 500 mg cholesterol in a small, similar
study showed that level of dietary cholesterol did not modify the
cholesterol-lowering effect of soy protein observed.
Descovich et al. (Ref. 33) examined the effects of soy protein
replacing animal protein in adults with stable type IIa and IIb
hypercholesterolemia. Subjects consumed an average of 47 g of soy
protein in the form of texturized soy protein (from SF) mixed into main
dishes. During the baseline control period with a lipid-lowering diet,
plasma total cholesterol decreased 3 percent from baseline levels. When
soy protein was substituted for animal protein in the second dietary
period, total cholesterol decreased by 24 percent (p<0.001) at the end
of the experimental period. All of the subjects demonstrated decreases
in total cholesterol of at least 10 percent. Upon returning to the
control diet, plasma total cholesterol increased 7 percent in men and 9
percent in women. LDL-cholesterol also showed a statistically
significant decrease, by 31 percent from baseline levels (p<0.001),
while HDL-cholesterol remained stable over the course of the soy
protein diet (+0.4 mg/dL for men and +1.0 mg/dL for women).
Wolfe et al. (Ref. 64) tested the effects of ingesting 47 g of soy
protein in the form of ISP incorporated into main dishes and a
beverage, while animal proteins were incorporated into similar main
dishes and cow's milk was consumed during the mixed protein control
period. Baseline lipid concentrations were not given; however, mean
total cholesterol concentrations were 280 mg/dL after the soy protein
treatment and 321 mg/dL after the control treatment. Thus, compared
with the control period, serum total cholesterol was 41 mg/dL lower
with ingestion of soy protein (p<0.05) and LDL-cholesterol was 43 mg/dL
lower (p<0.05). HDL-cholesterol was similar at the end of the soy
protein and control dietary periods.
Sirtori et al. (Ref. 56) conducted a trial that examined the
effects of complete and partial substitution of soy protein as SF (60 g
or 30 g of soy protein), in a lecithinated textured vegetable protein,
for animal protein in adults with type IIa hyperlipoproteinemia. Plasma
cholesterol levels were not altered during the first control diet
period. Total plasma cholesterol levels were significantly (p<0.01)
reduced in both periods of soy protein administration, by 18.6 percent
when 60 g were consumed and by 13.2 percent when 30 g were consumed.
Serum cholesterol values returned almost completely to baseline during
the second control period. Changes in LDL-cholesterol levels were
superimposable to those of total cholesterol. HDL-cholesterol levels
tended to increase during the two soy periods and decline to baseline
levels during the second control period, but these differences were not
statistically significant.
Verillo et al. (Ref. 60) compared the effects of substituting 31 g
of soy protein
[[Page 62986]]
as SF for animal protein versus the addition of 31 g of soy protein as
SF to animal protein in adults with stable type II
hypercholesterolemia. Slight, nonsignificant decreases in total and
LDL-cholesterol levels were reported during the initial control period.
Among subjects who consumed the soy-substituted diet, serum total
cholesterol declined significantly (p<0.01) from the end of the
baseline diet by 35 percent and 23 percent in type IIa and type IIb
patients, respectively. LDL-cholesterol declined significantly (p<0.01)
from the end of the baseline diet by 44 percent and 23 percent in type
IIa and type IIb patients, respectively. HDL-cholesterol increased 8
percent, but this change did not reach statistical significance. The
same hypocholesterolemic effects were also seen among subjects who
consumed the soy-added diet. A comparison of results at the ends of the
soy periods versus the means of final values of both control periods
showed differences in serum lipids that were of similar magnitudes, but
not statistically significantly different. The hypocholesterolemic
response to soy was significantly related to cholesterol level at entry
to the study.
The study of Lovati et al. (Ref. 46) in adults with type II
hypercholesterolemia provided soy protein as SF, from textured
vegetable protein, in amounts varying between 70 and 105 g depending
upon total energy consumed. Plasma total and LDL-cholesterol levels
both decreased by 16 percent (p<0.01) during the period when soy
protein diet was ingested compared with levels at the start of the
experimental period. Changes in these parameters on the control diet
were negligible. HDL-cholesterol concentrations were not documented but
were reported to be unchanged on the two diet regimens.
Gaddi et al. (Ref. 35) examined the effects of replacing animal
protein and non-soy plant protein with approximately 75 g soy protein
from SF in a lecithinated textured soy protein, in adults with familial
hypercholesterolemia. The control diet did not affect plasma lipid
values during the initial experimental period. After ingestion of the
soy protein diet, plasma total cholesterol decreased by 21 percent
(p<0.0l) and LDL-cholesterol decreased by 25 percent (p<0.01) from
levels measured after the first control diet period. HDL-cholesterol
levels were unchanged. Plasma total and LDL-cholesterol returned to
concentrations close to those at baseline following resumption of the
control diet during the third experimental period.
b. Studies in children. Gaddi et al. (Ref. 34) studied children
from 3 to 12 years of age with familial hypercholesterolemia. After a
baseline dietary period during which subjects consumed a low lipid
diet, soy protein in the form of SF replaced a portion of the animal
protein intake. No significant changes in plasma lipids occurred over
the duration of the baseline dietary period. Plasma total cholesterol
at the end of the soy protein dietary period was 20 percent lower than
at the end of the baseline dietary period (p<0.001). LDL-cholesterol
was 24 percent lower (p<0.01) and HDL-cholesterol level was not
affected.
Widhalm et al. (Ref. 63) examined the lipid-lowering effects of
incorporating ISP (13.5-18 g protein) into food and beverage recipes in
children with familial hypercholesterolemia. After the soy protein
dietary periods, plasma total cholesterol was 16 percent lower
(p<0.005) than baseline levels in the group that consumed the soy
protein diet before the control diet and 18 percent lower (p<0.001) in
the group that consumed soy last. LDL-cholesterol was also
statistically significantly decreased (p<0.001) by 22 percent in the
first group and 25 percent in the second group. During the control diet
periods, total and LDL-cholesterol levels were reduced by 8 percent and
7 percent in the first group and by 12 percent and 13 percent in the
second group, respectively. HDL-cholesterol was not statistically
significantly affected by dietary treatment.
Laurin et al. (Ref. 45) and Jacques et al. (Ref. 42) both reported
on a test of the lipid-lowering effects of ISP (28 g of soy protein) in
children, 6 to 12 years of age, with familial hypercholesterolemia.
Children consumed either a conventional low fat diet with 2-percent
cow's milk or the same low fat diet with a soy-based beverage made with
2-percent butterfat substituted for the 2-percent cow's milk.
Comparisons between the two treatment groups indicated that total and
LDL-cholesterol levels were not altered. HDL-cholesterol level was
increased 4 percent (p<0.04) with soy protein compared to cow's milk.
c. Summary--Subjects with Type II or familial hypercholesterolemia.
Each of the ten studies of the effects of soy protein in subjects with
severe (type II or familial) hypercholesterolemia employed diets low in
saturated fat and cholesterol (Refs. 33, 35, 46, 55, 56, 60, 64, 34, 42
and 45 (1 trial), and 63), and most subjects had been consuming such a
therapeutic diet prior to the study. Six of the ten trials were
conducted by workers from the same group (Refs. 55, 33, 56, 46, 35, and
34). Most used SF in TVP as the source of soy protein, in amounts
ranging from 14 to 105 g (Refs. 33, 56, 60, 46, 35, 34, and 63); the
remainder used ISP as the source of soy protein, in amounts ranging
from 28 to 60 g (Refs. 55, 64, and 42 and 45 (1 trial)). In all the
studies conducted in adults (Refs. 33, 35, 46, 55, 56, 60, and 64),
using both fixed sequence and crossover study designs, large and
statistically significant decreases in both total and LDL-cholesterol
levels were observed in response to consumption of diets containing soy
protein. In the six trials in which they were measured, HDL-cholesterol
levels were either not statistically significantly affected (Refs. 33,
64, 60, 46, and 35) or were statistically significantly increased (Ref.
56).
In the studies conducted in children with familial
hypercholesterolemia, two of the three trials demonstrated
statistically significant decreases from baseline levels in total and
LDL-cholesterol during the periods when soy protein diets were consumed
(Refs. 34 and 63). However, interpretation of these findings is
complicated by uncertainty about the control of intake of other dietary
constituents, especially saturated fat and cholesterol. In the study
reported by Laurin et al. and Jacques et al. (Refs. 45 and 42),
differences in these dietary components were controlled. With diets
providing 12 percent of kcal from saturated fat and 163 to 180 mg of
cholesterol, plasma total and LDL-cholesterol levels were not
statistically significantly different, but the HDL-cholesterol level
was statistically significantly higher, on the soy diet than on the
cow's milk diet.
3. Epidemiologic Evidence on Soy Protein and Blood Lipids
The petitioner also submitted one epidemiologic study by Nagata et
al. (Ref. 65) that described the relationship between soy product and
soy protein intake and serum total cholesterol concentrations in
Japanese men and women. Participants in this study were 1,242 men and
3,596 women from the Takayama Study, a prospective cohort study on the
impact of diet and lifestyle on cancer, who attended the annual health
checkup program between April and October 1992. Data regarding food
intake were collected by a validated, semiquantitative food frequency
questionnaire (FFQ). Blood samples were also taken for each subject and
analyzed for total cholesterol concentrations. Soy products identified
in the FFQ included tofu (plain, fried, deep-fried, or dried), miso,
fermented soybeans, soy milk, and boiled soybeans. The estimated amount
of soy protein consumed from these sources was 8.00 4.95
g/day for men and 6.88
[[Page 62987]]
4.06 g/day for women. The authors noted that their FFQ may
underestimate soy product intake; they also estimated that 4 to 9 g
additional soy protein may be consumed daily from soy protein added to
meats and fish pastes that was not accounted for in the FFQ. Thus,
analyses were presented in terms of relative soy protein intake. Using
energy-adjusted means for quartiles of soy protein intake, a
statistically significant negative trend was observed for lower serum
total cholesterol concentrations with higher levels of soy protein
intake (p<0.0001 for both men and women). The analysis for men was
controlled for age, smoking status, and total energy, protein, and fat
intake. The analysis for women was controlled for age, menopausal
status, body mass index, and intake of energy and vitamin C. Further
adjustments for physical activity, coffee and tea consumption, and
intakes of cholesterol, carbohydrates, fiber, and vitamin E were
performed and results were not affected. Between the 1st and 4th
quartiles in men, total cholesterol was lower by 12 mg/dL with a 9.6-g
increase in soy protein intake. For women, total cholesterol was lower
by 9 mg/dL with a 7.9-g increase in soy protein intake.
4. Meta-analysis of Studies of Soy Protein and Blood Lipids
The petitioner presented the results of a 1995 meta-analysis (Ref.
66) of the effect of soy protein on blood lipids. While the role of
``research synthesis'' studies, including meta-analyses, is of
interest, it is as yet unresolved. The appropriateness of such
analytical techniques to establish diet/health relationships in
particular is not known. There are on-going efforts to identify
criteria and critical factors to consider in both conducting and using
such analyses, but this science is still emerging. Therefore, the meta-
analysis did not weigh heavily within the body of evidence for this
relationship.
In summary, Anderson et al. (Ref. 66) pooled data from studies that
were deemed comparable in methodology in order to perform a meta-
analysis of the effect of soy protein on blood lipids. Of the 37
publications identified by these investigators that presented data on
soy protein and lipid changes, 29 met the criteria of using either ISP
or texturized soy protein as the soy protein source, employing either a
parallel or crossover design, and providing initial or baseline
cholesterol values to allow calculation of decreases. These 29
publications reported the findings from 38 separate trials. Each of
these publications was included in the petition and was considered for
review individually by FDA as described previously. Thirty-four of the
trials were conducted among adults and four among children. Study
samples included individuals with normal blood cholesterol levels as
well as those with mildly to severely elevated levels. Twelve of the
trials were conducted in subjects with familial hypercholesterolemia.
The specific analytical approach is described in Anderson et al.
(66). Based on examining the difference from baseline between the soy
protein and control protein groups, the analysis indicated that soy
protein consumption statistically significantly decreased total
cholesterol for the pooled data by 9.3 percent and LDL-cholesterol by
12.9 percent. HDL-cholesterol was increased by a net of 2.4 percent
with soy protein ingestion, but this change was not statistically
significant. This analysis also suggested that the initial level of
serum total cholesterol was the most important determinant of serum
lipid response to soy protein. When changes in total and LDL-
cholesterol were examined by quartile of baseline cholesterol
concentration, a progressively greater magnitude of change was observed
from the lowest to the highest quartiles. Additional analyses indicated
that the type and amount of soy protein consumed and type of background
diet did not substantially influence the results.
To examine further the effects of the type and amount of soy
protein, an analysis was performed using changes observed with the soy
diet alone instead of net changes as the outcome variable. Initial
serum cholesterol concentration was also the major determinant of
effects in this model, but statistically significant effects (p=0.02)
were also obtained for amount of soy protein. This model predicted,
after adjustment for initial values and other variables, serum total
cholesterol decreases of 8.9 mg/dL with 25 g/day soy protein, 17.4 mg/
dL with 50 g/day of soy protein, and 26.3 mg/dL with 75 g/day of soy
protein.
5. Studies of the Role Soy Isoflavones
Isoflavones are a class of naturally-occurring compounds with weak
estrogenic/antiestrogenic activities that are present in a wide variety
of plants. The 12 major isomers of naturally-occurring isoflavones in
soybeans are genistein, genestin, 6``-O-acetylgenistin, 6``-O-
malonylgenestin, diadzein, diadzin, 6``-O-acetyldiadzin, 6``-O-
malonyldiadzin, glycitein, glycitin, 6``-O-acetylglycitin, and 6``-O-
malonyglycitin. The levels of isoflavones in soybeans are known to vary
with cultivar and growing conditions. Soy isoflavones are retained to
variable degrees in soy protein products and soy foods, depending on
the particular processing techniques used. For example, essentially all
of the isoflavones in soy protein products can be extracted by alcohol
washing, and their levels can also be reduced by repeated aqueous
washings and some texturization techniques. Because of the estrogenic
activities of the soy isoflavones, they have been hypothesized to
contribute to the hypocholesterolemic effect of soy protein.
The petitioner submitted an unpublished study by Crouse et al.
(Ref. 31, documented in Ref. 1 with corrections noted in Ref. 2) that
examined the effect of soy protein containing different levels of
isoflavones in hypercholesterolemic men and women (summarized in Table
1 of this document). Potential subjects were provided instruction in an
NCEP Step 1 diet and followed this diet for 1 month. Subjects with
qualifying serum lipid levels (LDL-cholesterol >140 mg/dL) were given a
casein drink containing 25 g protein to consume in place of other
protein in the NCEP Step 1 diet. Subjects compliant with this regimen
were then randomized into one of five treatment groups and baseline
blood lipid values were obtained. The treatment groups received 25 g
protein from ISP prepared from soy with different levels of isoflavones
(either 1.0, 1.6, or 2.5 mg total aglycone isoflavones/g protein), or
25 g protein from alcohol-washed ISP that contained essentially no
isoflavones (0.2 mg total aglycone isoflavones/g protein) or 25 g
protein from casein (no isoflavones) in beverages for 9 weeks. Dietary
intake was assessed at baseline and at the end of the study. Diet was
reported to be stable and comparable between groups throughout the
study, with 9 percent of energy derived from saturated fat. Body weight
was also stable, with no differences between groups at baseline or at
the end of the trial. Results indicated the ISP containing the highest
level of isoflavones significantly lowered total (p<0.05) and LDL-
cholesterol (p<0.05), by 4 percent and 6 percent, respectively, while
HDL-cholesterol was not altered (Table 1). Furthermore, in subjects
with LDL-cholesterol in the top half of the study population, serum
total and LDL-cholesterol were reduced by 9 percent (p<0.03) and 12
percent (p<0.03), respectively, by the ISP with the highest isoflavone
content, and by 8 percent (p<0.03) and 9 percent (p<0.03), respectively
by the ISP with the second highest isoflavone content, while HDL-
cholesterol concentrations were maintained.
[[Page 62988]]
Baum et al. (Ref. 28) also investigated the impact in soy protein
containing different levels of isoflavones on cholesterol lowering and
examined whether changes in blood lipids were lasting or transient.
Subjects were moderately hypercholesterolemic women, who were at least
1 year since last menstrual period, and were not taking medications
known to alter lipid or bone metabolism. Following a 2-week run-in
period during which subjects consumed an NCEP Step I diet, subjects
were randomly assigned to one of three treatment groups consisting of
40 g protein from either ISP with 1.4 mg total aglycone isoflavones/g
protein, ISP with 2.3 mg total aglycone isoflavones/g protein, or
casein/nonfat dry milk for the 24-week treatment period. Although the
adjusted mean difference in total serum cholesterol level was not
statistically significantly different in the soy groups and the control
group, there was a significant reduction of 8 to 9 percent in non-HDL
(LDL plus VLDL) cholesterol in both of the ISP treatment groups
(p=0.04) compared to the control group. HDL-cholesterol was also
significantly increased in both soy groups compared to the control.
Body weight remained stable, and dietary intake was assessed and was
reported to be similar among treatment groups although details were not
reported.
The petitioner concluded that these two studies (Refs. 31 and 28)
provided evidence that the hypocholesterolemic effect of soy protein is
dependent on processing techniques that enable retention of the
naturally occurring isoflavones in conjunction with the soy protein. As
additional supportive evidence for this conclusion, the petitioner
cited studies of the lipid-lowering effects of soy protein with
naturally occurring isoflavones in nonhuman primates (Refs. 22 and 70).
In these experiments, the effects of diets including ISP with naturally
occurring isoflavones compared with those of diets containing either
casein or alcohol-washed ISP stripped of essentially all naturally
occurring isoflavones were examined in two species of monkeys. The
studies demonstrate significant depressions in total and non-HDL (LDL
plus VLDL) cholesterol levels in response to diets containing
unextracted ISP as compared with the diets containing casein or
alcohol-washed ISP. As evidence that soy isoflavones alone, in the
absence of soy protein, are ineffective in lowering blood lipids, the
petitioner cited the study of Nestel et al. (Ref. 71). In that study,
consumption of a tablet containing 80 mg of total aglycone isoflavones
(mainly genistein and diadzein) had no impact on blood lipid profiles
in postmenopausal women.
Although the petitioner suggested, based on the studies of Crouse
et al. (Ref. 31) and Baum et al. (Ref. 28), that isoflavone content
exceeding a certain threshold was a useful marker for soy protein that
would be effective in lowering blood lipid levels, FDA has tentatively
concluded that the evidence is not sufficient to establish that the
presence of isoflavones accounts for or is related to the effect on
blood lipids. The agency notes that there are a variety of methods for
processing soy that could give rise to variable amounts of naturally-
occurring isoflavones in soy protein products, and this is a possible
hypothesis for explaining some of the variability in the results of
human intervention studies. However, with two exceptions (Refs. 31 and
28), the studies reviewed and described in this document did not
include concurrent measures of the isoflavone content of the soy
protein products studied. More importantly, a recent letter to the
editor from Sirtori et al. (Ref. 72), which was not included in the
petition, contradicts the conclusions of Crouse et al. (Ref. 31) and
Baum et al. (Ref. 28). These researchers (Ref. 72) reported that the
TVP fed in their studies contained essentially no isoflavones and still
considerable impact on LDL-cholesterol was observed. These studies
(Refs. 33, 56, 46, 35, and 34) were conducted in subjects with type II
hypercholesterolemia and all showed large and significant decreases in
blood total and LDL-cholesterol levels.
Given the limited number of studies and the contradictory outcomes,
FDA is not persuaded that the isoflavone component of soy protein is a
relevant factor to the diet-disease relationship. Rather, FDA
tentatively concludes that the evidence from a wide range of studies
using differently processed soy protein is supportive of a relationship
between soy protein per se and reduced risk of CHD.
6. Summary
In five (Refs. 31, 28, 27, 51, and 44) of seven (Refs. 31, 28, 27,
51, 44, and 40 (2 trials)) well-controlled studies of
hypercholesterolemic subjects consuming low saturated fat and low
cholesterol diets, soy protein intake was associated with statistically
significant decreases in total and/or LDL-cholesterol levels, either in
the entire study populations or subsets of subjects with higher initial
blood lipid levels. Levels of HDL-cholesterol were found either to be
unchanged (Refs. 31, 27, 51, and 40 (2 trials)) or slightly but
statistically significantly increased (Refs. 28 and 44) by consumption
of diets containing soy protein.
Levels of soy protein as ISP found to be effective in lowering
total and LDL-cholesterol levels, in the context of a diet low in
saturated fat and cholesterol, ranged in these studies from 25 to 50 g
(Refs. 31, 28, 27, and 51). As whole soybean milk, 31 g of soy protein
lowered LDL-cholesterol only in the subset of subjects with the highest
initial LDL-cholesterol levels and LDL/HDL-cholesterol levels (Ref.
44). Diets providing 50 g of soy protein as SF did not cause
significant changes from baseline for any of the blood lipids, but the
decrease in total cholesterol observed after ingestion of SF was
significantly greater than that on the control diet of nonfat dry milk
and cellulose (Ref. 51). Diets providing 27 g of soy protein as SF in a
textured product had no significant effects on blood lipid levels
compared to a control diet, and a higher level (62 g) significantly
lowered total cholesterol only in the experimental group fed the soy
protein diet first (Ref. 40).
Three intervention studies reported on effects of soy protein in
hypercholesterolemic subjects consuming ``usual'' diets that were
generally high in total fat, saturated fat, and cholesterol (Refs. 37,
49, and 54). In each of these three studies, soy protein was fed in
experimental diets as ISP (Refs. 37, 49, and 54). With a diet
containing a very high level (99 g) of soy protein from this source
(Ref. 37), statistically significant differences in both total and LDL-
cholesterol were reported. Results were less consistent, showing a
significant decrease in total cholesterol only in subjects with the
highest baseline levels, with a relatively low level of soy protein (17
g) (Ref. 49). An intermediate level of soy protein (57 g) was found to
be ineffective in lowering total cholesterol in the study of Shorey et
al. (Ref. 54). (This latter study had a number of anomalous results.)
HDL-cholesterol was also measured in these three studies. Changes were
small and not statistically significant in two studies (Refs. 37 and
49), but HDL-cholesterol was statistically significantly decreased from
baseline values by 15 percent in one study (Ref. 54).
Five intervention studies examined the effects of soy protein in
normocholesterolemic subjects (Refs. 30, 36, 58, 59, and 53). Effects
of soy protein on total and LDL-cholesterol were less consistent in
normocholesterolemic subjects than in hypercholesterolemic subjects. As
ISP, 44 g of soy protein was effective in significantly lowering total
cholesterol in one study (Ref. 30) and 54 g
[[Page 62989]]
significantly lowered LDL-cholesterol in one study (Ref. 58). With very
low initial blood lipid levels seen in some of these studies, the
impact of dietary changes is considerably lessened. HDL-cholesterol was
measured in four of these studies, with statistically significant
increases associated with soy protein intake found in two (Refs. 58 and
59) and no statistically significant changes in two (Refs. 36 and 53).
Two intervention studies (Refs. 29 and 43) examined the
hypocholesterolemic effects of soy protein versus casein in the context
of hypocaloric diets fed to obese persons to achieve significant weight
reduction. These two studies (Refs. 29 and 43) demonstrated large
decreases in both total and LDL-cholesterol levels during hypocaloric
diets that provided relatively low amounts (27 and 17 g, respectively)
of soy protein. HDL-cholesterol was decreased in both soy and casein
groups at the end of the first study (Ref. 29); however, only the
casein group's values were significantly different (p<0.01) from
baseline. Additionally, the decrease in HDL-cholesterol in the casein
group was significantly greater (p<0.01) than that observed in the soy
protein group. In the second study (Ref. 43), HDL-cholesterol levels
were not significantly affected by dietary treatment.
In all seven intervention studies conducted in adults with type II
or familial hypercholesterolemia (Refs. 33, 35, 46, 55, 56, 60, and
64), large and statistically significant decreases in both total and
LDL-cholesterol levels were observed in response to consumption of
diets containing soy protein. In the six trials in which they were
measured, HDL-cholesterol levels were either not statistically
significantly affected (Refs. 33, 64, 60, 46, and 35) or statistically
significantly increased (Ref. 56). Each of these studies in adults with
severe (type II or familial) hypercholesterolemia employed diets low in
saturated fat and cholesterol (Refs. 33, 35, 46, 55, 56, 60, and 64)
and most subjects had been consuming such a therapeutic diet prior to
the study. Most trials used SF in TVP as the source of soy protein, in
amounts ranging from 31 to 105 g (Refs. 33, 56, 60, 46, and 35); the
remainder used ISP as the source of soy protein, in amounts ranging
from 28 to 60 g (Refs. 55 and 64). Two of the three trials conducted in
children with familial hypercholesterolemia demonstrated significant
decreases from baseline levels in total and LDL-cholesterol during the
periods when soy protein diets were consumed (Refs. 34 and 63).
Evidence from one epidemiologic study (Ref. 65) supported a
significant negative trend for lower serum total cholesterol
concentrations with higher levels of soy protein intake (p<0.0001 for
both men and women). Between the first and fourth quartiles in men
total cholesterol was lower by12 mg/dL with a 9.6-g increase in soy
protein intake. For women, total cholesterol was lower by 9 mg/dL with
a 7.9-g increase in soy protein intake.
Based on these studies, FDA tentatively finds there is scientific
evidence for a consistent, clinically significant effect of soy protein
on blood total and LDL-cholesterol. The hypocholesterolemic effect of
soy protein is seen in addition to the effects of a low saturated fat
and low cholesterol diet. The degree of lowering of blood total and
LDL-cholesterol is consistently and highly dependent on initial levels,
within and across studies of subjects with normal, moderately elevated,
and severely elevated blood lipid levels, with persons having higher
blood lipid levels showing greater effects. Soy protein consistently
causes only statistically nonsignificant effects or slight elevations
in HDL-cholesterol levels. The intervention studies suggest that a
minimum level of approximately 25 g of soy protein is needed to have a
clinically significant effect on total and LDL-cholesterol levels.
These conclusions, drawn from the review of the individual, well
controlled studies, are also supported by the meta-analysis of Anderson
et al. (66).
IV. Decision To Propose a Health Claim Relating Soy Protein to
Reduction in Risk of CHD
The petition provided and FDA reviewed information on pertinent
human studies that evaluated the effects on serum cholesterol and LDL-
cholesterol levels from dietary intervention with soy protein in
subjects with normal to elevated serum cholesterol levels.
FDA tentatively concludes that, based on the totality of publicly
available scientific evidence, there is significant scientific
agreement to support the relationship between consumption of soy
protein included in a diet low in saturated fat and cholesterol and the
risk of CHD. The strongest evidence for the effect of soy protein on
the risk of CHD is provided by studies that measured the effect of
dietary soy protein consumption on the two major risk factors for CHD,
total and LDL-cholesterol.
In most intervention trials in subjects with total cholesterol<300
mg/dL, soy protein was found to reduce total and/or LDL-cholesterol
levels to a clinically significant degree (Refs. 31, 28, 27, 51, 44,
37, 49, 30, 58, 29, and 43). Moreover, HDL-cholesterol levels were
unchanged (Refs. 31, 27, 51, 40, 37, 49, 36, and 53) or slightly
increased (Refs. 28, 44, 58, and 59). In some cases (Refs. 27, 44, and
49), decreases in total and LDL-cholesterol were statistically
significant only in subsets of subjects with the higher initial blood
lipid levels. Results in normocholesterolemic subjects (Refs. 30, 36,
58, 59, and 53) were more variable than in hypercholesterolemic
subjects (31, 28, 27, 51, 44, 40, 37, 49, 54, 29, and 43). The outcome
of an epidemiologic study (Ref. 65) also supported a relationship
between higher levels of soy protein intake and lower blood lipid
levels.
Most of the studies in subjects with total cholesterol<300 mg/dL
used low saturated fat and low cholesterol diets (Refs. 31, 28, 27, 51,
44, 30, 36, 53, 29, and 43), but some used ``usual'' diets (Refs. 37,
49, 54, 36, 58, and 59). Although soy protein was found to lower blood
lipid levels in some of the studies using ``usual'' diets,
hypocholesterolemic effects of soy protein were more consistently
observed with diets low in saturated fat and cholesterol. In some
studies (especially those without run-in periods) (Refs. 40 and 54),
the control low saturated fat and low cholesterol diets induced
significant decreases in blood lipid levels making it difficult to
detect any additional effect of soy protein. At the same time, in two
studies in which soy protein containing hypocaloric diets were compared
to similar diets without soy (Refs. 29 and 43), only the soy protein
containing diets induced significant changes in blood lipid levels.
Given the variability of amounts and forms in which soy protein was
provided in the diets, the response of blood lipid levels appears
robust.
Data from studies of adults with type II and familial forms of
hypercholesterolemia (and total cholesterol levels in excess of 300 mg/
dL) (Refs. 55, 33, 64, 56, 64, 46, and 35) were more consistent than
studies in persons with lower blood lipid levels in showing large and
statistically significant decreases in total and LDL-cholesterol,
accompanied by no changes or slight increases in HDL-cholesterol
levels. Nearly all of the subjects in these trials consumed low
saturated fat and low cholesterol diets during the studies and had
consumed such diets prior to studies with soy protein.
Soy protein was tested in a variety of food forms (as soy
beverages, formulated into meat and dairy product analogs, added to
soups, or baked into foods, such as muffins and breads) but produced
fairly consistent results
[[Page 62990]]
regardless of the food form fed and apparent differences in processing
techniques.
FDA tentatively concludes, based on the evidence submitted and
reviewed, that soy protein, included in a diet low in saturated fat and
cholesterol, can lower blood total and LDL-cholesterol levels, without
adversely affecting HDL-cholesterol levels. The agency also tentatively
concludes that the effect is due to soy protein per se and is not
consistently related to the presence or absence of isoflavones. The
intervention studies suggest that a minimum level of approximately 25 g
of soy protein is needed to have a clinically significant effect on
total and LDL-cholesterol levels.
Based on the totality of the scientific evidence presented in the
petition, the agency tentatively concludes that there is significant
scientific evidence to show that soy protein, included in a diet low in
saturated fat and cholesterol, will help reduce serum lipids, and that
such reductions may reduce the risk of CHD. In the majority of clinical
studies evaluating soy products, total and LDL-cholesterol were the
lipid fractions shown to be the most affected by soy protein
intervention. As part of a diet low in saturated fat and cholesterol,
regular consumption of soy protein, in an amount to provide 25 g/day,
resulted in reduced total and LDL-cholesterol levels in subjects with
normal and elevated serum cholesterol levels. As stated in section
III.A of this document, Federal Government and other reviews have
concluded that there is substantial epidemiologic and clinical evidence
that high blood levels of total cholesterol and LDL-cholesterol
represent major contributors to CHD (56 FR 60727 at 60728, and Refs. 4
through 7). Dietary factors that decrease total cholesterol and LDL-
cholesterol will affect the risk of CHD (Refs. 4 through 7).
Given all of this evidence, the agency is proposing a health claim
on the relationship between soy protein and reduced risk of CHD.
V. Description and Rationale for Components of Health Claim
A. Relationship Between Soy Protein and CHD and the Significance of the
Relationship
Proposed Sec. 101.82(a) describes the relationship between diets
low in saturated fat and cholesterol containing soy protein and the
risk of CHD. In proposed Sec. 101.82(a)(1), the agency recounts that
CHD is the most common and serious form of CVD, and that CHD refers to
diseases of the heart muscle and supporting blood vessels. The proposed
section also notes that high blood total and LDL-cholesterol levels are
associated with increased risk of developing CHD and identifies the
levels of total cholesterol and LDL-cholesterol that would put an
individual at high risk of developing CHD, as well as those serum lipid
levels that are associated with borderline high risk. This information
will assist consumers in understanding the seriousness of CHD.
In proposed Sec. 101.82(a)(2), the agency recounts that populations
with a low incidence of CHD tend to have low blood total and LDL-
cholesterol levels. It states that these populations also tend to have
dietary patterns that are low in total fat, saturated fat, and
cholesterol and high in plant foods that contain fiber and other
components. This information is consistent with that provided in the
authorized health claim for fruits, vegetables, and grain products and
CHD (Sec. 101.77) and so the agency believes that this information
provides a basis for a better understanding of the numerous factors
that contribute to the risk of CHD and the relationship with soy
protein and diets low in saturated fat and cholesterol.
Proposed Sec. 101.82(a)(3) states that diets low in saturated fat
and cholesteral may reduce the risk of CHD. The paragraph further
states that soy protein, when added to such a diet, may also help
reduce the risk of CHD.
Proposed Sec. 101.82(b) describes the significance of the diet-
disease relationship. In proposed Sec. 101.82(b)(1), the agency
recounts that CHD remains a major public health concern in the United
States because the disease accounts for more deaths than any other
disease or group of diseases. The claim states that early management of
modifiable risk factors for CHD is a major public health goal that can
assist in reducing the risk of CHD. This information is consistent with
the evidence that lowering blood total and LDL-cholesterol levels
reduces the risk of CHD (56 FR 60727, 58 FR 2739, and Refs. 4 through
8).
In proposed Sec. 101.82(b)(2), the significance of the relationship
between soy protein and CHD risk factors in context of the total diet
is discussed. The agency recounts that many Americans' intakes of
saturated fat and cholesterol exceed recommended levels, and it
summarizes public health recommendations for the diet (56 FR 60727 at
60738 and Sec. 101.75(b)(3)). This paragraph also states that
scientific evidence demonstrates that diets low in saturated fat and
cholesterol and that contain soy protein are associated with reduced
blood lipids. FDA tentatively concludes that the latter statement is
scientifically valid based on the evidence that it has reviewed on this
nutrient-disease relationship.
B. Nature of the Claim
In proposed Sec. 101.82(c)(1), FDA is proposing to require that all
of the general requirements for health claims set out in Sec. 101.14 be
met. This provision is consistent with the provisions of the other
specific health claim regulations in 21 CFR part 101, subpart E.
In proposed Sec. 101.82(c)(2)(i), FDA is proposing to authorize a
health claim on the relationship between diets that contain soy protein
and are low in saturated fat and cholesterol and the risk of CHD. The
agency is proposing to do so based on its review of the scientific
evidence on this nutrient-disease relationship, which shows that diets
that contain soy protein and are low in saturated fat and cholesterol
help to reduce total and LDL-cholesterol levels, especially in
individuals with elevated blood total cholesterol (Refs. 31, 28, 27,
51, 44, 37, 49, 30, 58, 29, 43, 55, 33, 64, 56, 64, 46, and 35). This
result is significant for the risk of heart disease because elevated
levels of total and LDL-cholesterol are associated with increased risk
of CHD (Refs. 4 through 7).
Most of the scientific evidence for an effect of soy protein on
blood lipid levels was provided by studies that used diets low in
saturated fat and cholesterol. Although soy protein was found to lower
blood lipid levels in some of the studies using ``usual'' diets (Refs.
37, 49, and 58), hypocholesterolemic effects of soy protein were more
consistently observed with diets low in saturated fat and cholesterol.
Moreover, as stated in section V.A of this document, CHD is a major
public health concern in the United States, and the totality of the
scientific evidence provides strong and consistent support that diets
high in saturated fat and cholesterol are associated with elevated
levels of blood total and LDL-cholesterol and, thus, CHD (56 FR 60727
at 60737). Dietary estimates for American adults show that the average
saturated fat intakes of American adults are about 13 percent of
calories, total fat intakes are about 37 percent of calories, and
average cholesterol intakes range from 300 to over 400 mg daily for
adult men and women (56 FR 60727 at 60738). The current intakes of
saturated fat and total fat are thus well in excess of recommended
goals of less than 10 percent and 30 percent of calories.
[[Page 62991]]
Dietary guidelines from both Government and private scientific bodies
conclude that the majority of the American population would benefit
from decreased consumption of dietary saturated fat and cholesterol
(Refs. 4 through 7). The results of several studies showed that daily
consumption of soy protein lowered total cholesterol and LDL-
cholesterol, and the effects of dietary intake of soy protein were
evident when the diets were low in saturated fat and cholesterol (Refs.
31, 28, 27, 51, 44, 30, 29, and 43). Thus, the agency tentatively finds
that it will be more helpful to Americans' efforts to maintain healthy
dietary practices if the effect of soy protein on serum lipids is
characterized in the context of a diet low in saturated fat and
cholesterol.
In Sec. 101.82(c)(2)(i)(A), the agency is proposing to require,
consistent with other health claims, that the relationship be qualified
with the terms ``may'' or ``might.'' These terms are used to make clear
that not all persons can necessarily expect to benefit from these
dietary changes (56 FR 60727 at 60740 and 58 FR 2552 at 2573). The
requirement that the claim use the term ``may'' or ``might'' to relate
the ability of soy protein to reduce the risk of heart disease is also
intended to reflect the multifactorial nature of the disease.
In Sec. 101.82(c)(2)(i)(B), the agency is proposing to require,
consistent with other authorized health claims, that the terms
``coronary heart disease'' or ``heart disease'' be used in specifying
the disease. These terms are commonly used in dietary guidance
materials, and therefore they should be readily understandable to the
consumer (56 FR 60727 at 60740 and 58 FR 2552 at 2573).
In Sec. 101.82(c)(2)(i)(C), the agency is proposing that the claim
specify the substance as ``soy protein.'' Based on its review of the
scientific evidence submitted with the petition, the agency tentatively
concludes that there is significant scientific agreement that diets low
in saturated fat and cholesterol that contain soy protein may help to
reduce blood total and LDL-cholesterol levels, the major modifiable
risk factors for CHD (Refs. 31, 28, 27, 51, 44, 37, 49, 30, 58, 29, 43,
55, 33, 64, 56, 64, 46, and 35). As discussed in section III.C.5 of
this document, FDA did not find persuasive the limited and
contradictory evidence that soy isoflavones are a relevant factor in
the diet-disease relationship persuasive. Therefore, FDA has
tentatively concluded that evidence from a wide range of studies
supports a relationship between soy protein per se and reduced risk of
CHD.
As discussed previously, the agency tentatively finds that for the
public to understand fully, in the context of the total daily diet, the
significance of consumption of soy protein on the risk of CHD (see
section 403(r)(3)(B)(iii) of the act), information about the total diet
must be included as part of the claim. Therefore, in
Sec. 101.82(c)(2)(i)(D), the agency is proposing to require that the
claim include the fact that the effect of dietary consumption of soy
protein on the risk of CHD is evident when it is consumed as part of a
healthy diet and that, consistent with other authorized health claims,
the fat component of the diet be specified as ``saturated fat'' and
``cholesterol.'' Based on its review of the scientific evidence
submitted with the petition, the agency tentatively concludes that
there is significant scientific agreement that diets containing soy
protein and low in saturated fat and cholesterol are associated with
reduced blood total and LDL-cholesterol levels.
Proposed Sec. 101.82(c)(2)(i)(E), consistent with other authorized
health claims, requires that the claim not attribute any degree of risk
reduction of CHD to consumption of diets low in saturated fat and
cholesterol that contain soy protein. None of the studies that the
agency reviewed provide a basis for determining the percent reduction
in risk of CHD likely from consuming diets that contain soy protein and
are low in saturated fat and cholesterol. Also consistent with other
authorized claims, proposed Sec. 101.82(c)(2)(i)(F) requires that the
claim not imply that consumption of diets low in saturated fat and
cholesterol and that contain soy protein is the only recognized means
of reducing CHD risk.
Proposed Sec. 101.82(c)(2)(i)(G) requires that the claim specify
the daily dietary intake of soy protein needed to reduce the risk of
CHD and the contribution one serving of the product makes to achieving
the specified daily dietary intake. This requirement is consistent with
requirements set forth in Sec. 101.81.
In the studies showing a statistically significant effect of soy
protein on total or LDL-cholesterol, the amounts fed ranged from 17 to
105 g/day (Refs. 31, 28, 27, 51, 44, 37, 49, 30, 58, 29, 43, 55, 33,
64, 56, 64, 46, and 35). In proposing 25 g/day as an effective daily
intake of soy protein, the petitioner relied on the meta-analysis by
Anderson et al. (Ref. 65) and noted that the estimate suggested by the
meta-analysis was confirmed by the recent study of Crouse et al. (Ref.
31) that found reductions in total and LDL-cholesterol of 4 and 6
percent, respectively, with ingestion of 25 g/day of soy protein
containing high levels of isoflavones.
FDA notes that, although none of the studies reviewed attempted to
determine an effective or optimal amount of soy protein, the study by
Sirtori et al. (Ref. 56) suggests the existence of a dose-response. In
that study of subjects with type II hypercholesterolemia, total
cholesterol levels were reduced by 13 and 19 percent, and LDL-
cholesterol levels were reduced by 18 and 23 percent, compared to
control by ingestion of 30 and 60 g/day of soy protein, respectively.
With levels of soy protein intake lower than the proposed effective
amount, findings have been variable. Mercer et al. (Ref. 49) found a
statistically significant reduction in total cholesterol in response to
17 g/day of soy protein only in those subjects with high initial
values. Feeding the same amount (17 g/day) of soy protein in a
hypocaloric diet, however, Jenkins et al. (Ref. 43) found statistically
significant reductions of 10 and 17 percent in total and LDL-
cholesterol, respectively. With 25 g/day of soy protein, Bakhit et al.
(Ref. 27) found a statistically significant reduction in total
cholesterol (about 8 percent) in subjects with blood cholesterol levels
greater than 220 mg/dL. Crouse et al. (Ref. 31) found that 25 g of soy
protein that contained a high level of isoflavones significantly
lowered total (p<0.05) and LDL-cholesterol (p<0.05), by 4 percent and 6
percent, respectively. Furthermore, in subjects with LDL-cholesterol in
the top half of the study population, serum total and LDL-cholesterol
were reduced by 9 percent (p<0.03) and 12 percent (p<0.03),
respectively, by soy protein with the highest isoflavone content, and
by 8 percent (p<0.03) and 9 percent (p<0.03), respectively, by soy
protein with the second highest isoflavone content. Although Holmes et
al. (Ref. 40) did not find statistically significant changes in blood
lipids with 27 g of soy protein, using 28 g of soy protein in a
hypocaloric diet, Bosello et al. (Ref. 29) observed decreases of 16
percent from baseline in both total and LDL-cholesterol (p<0.01). With
31 g of soy protein, Kurowska et al. (Ref. 44) found an 11-percent
reduction in LDL-cholesterol in subjects with the highest initial LDL-
cholesterol levels and LDL/HDL-cholesterol ratios. As a substitution or
as an addition, Verillo et al. (Ref. 60) found 31 g of soy protein
produced large (>20 percent) reductions in both total and LDL-
cholesterol in subjects with type II hypercholesterolemia.
Based on these data that support a dose-response and that show
clinically significant reductions in total and LDL-
[[Page 62992]]
cholesterol with soy protein ingestion in the range of 17 to 31 g/day,
and recognizing that the hypocholesterolemic effects of soy protein are
highly dependent on initial blood lipid levels, the agency has
tentatively accepted that 25 g/day represents a reasonable effective
amount of soy protein. In addition, an amount of 25 g/day of soy
protein represents half of the Reference Daily Intake (RDI) of 50 g for
protein and is a reasonable level of consumption in the context of the
total daily diet. Thus, FDA tentatively concludes that the amount of
soy protein associated with reduction in total and LDL-cholesterol
levels and, thus, with reduced risk of CHD is 25 g or more of soy
protein per day. The agency is asking for comments on this tentative
determination.
C. Nature of the Substance
Proposed Sec. 101.82(c)(2)(ii)(A) indicates that soy protein from
the legume seed Glycine max is the substance that is the subject of
this claim.
Proposed Sec. 101.82(c)(2)(ii) (B) sets out FDA's tentative
decision that soy protein when evaluated for compliance purposes by the
agency will be measured using the Association of Official Analytical
Chemists International (AOAC) official method of analysis No. 988.10.
The petitioner proposed that measurement of total soy isoflavones
be used as a marker for the content of soy protein in foods and as an
indicator of the effectiveness of soy protein products in reducing
blood cholesterol. As discussed in section C.III.5 of this document,
FDA disagrees with the petitioner's conclusions regarding the
significance of soy isoflavones with respect to the observed
hypocholesterolemic effects of soy protein. Accordingly, FDA finds the
proposed methodology to assess isoflavones irrelevant. The AOAC method
that FDA is proposing instead is an enzyme-linked immunosorbant assay
that detects soy protein in raw and heat-processed meat products. With
this assay, samples are compared to standard commercial soy protein and
appropriate blanks. The method is described as semi-quantitative, but
it can be quantitative when the nature of the soy protein in the
samples is known and the assay is calibrated accordingly. The sample
extraction procedure, which involves preparation of an acetone powder,
has been shown to be appropriate for a complex food matrix (meat). FDA
believes, therefore, that this assay should also be suitable for other
food matrices. FDA is requesting comments on the suitability of this
method for assuring that foods bearing the health claim contain
qualifying levels of soy protein.
D. Nature of the Food
Proposed Sec. 101.82(c)(2)(iii)(A) requires that the food bearing
the health claim contain at least 6.25 g of soy protein per reference
amount customarily consumed (RACC) of the food product.
Using 25 g of soy protein as the qualifying amount for a CHD claim,
the petitioner suggested that a single serving of a soy protein-
containing product (i.e., 1 RACC) should provide 1/4 of this amount
(based on 4 servings a day). Thus, a soy protein-containing product
would have to contain at least 6.25 g soy protein (1/4 x 25 g) per
RACC. The petitioner stated that this approach is reasonable because it
would permit a wide variety of low fat, soy protein containing products
to bear the health claim. The petitioner provided a list of products on
the market that currently meet the proposed requirements and a list of
products that could be modified to meet them (Ref. 1, Appendix V).
The agency has generally made the assumption that a daily food
consumption pattern includes three meals and a snack (see 58 FR 2302 at
2379, January 6, 1993). Because of the wide variety of types of foods
that could contain qualifying levels of soy protein (baked goods, tofu,
soy beverages and shakes, meat analogs), the agency has tentatively
concluded that the assumption of 4 servings/d of soy protein containing
foods is reasonable. Therefore, the agency tentatively finds that use
of the qualifying criterion set forth in the petition for this proposed
rule is appropriate and is proposing this level in this document. The
qualifying level of protein, 6.25 g/RACC, exceeds the amount required
for a food to qualify as a ``good source'' of protein, i.e., 10 percent
of the RDI of 50 g or 5 g/RACC).
In Sec. 101.82(c)(2)(iii)(B), the agency is proposing, consistent
with other authorized heart disease health claims, that foods bearing
the health claim meet requirements for ``low saturated fat,'' ``low
cholesterol,'' and ``low fat.'' In the preamble to the final rule on
fruits, vegetables, and grain products and heart disease (Sec. 101.77,
58 FR 2552 at 2572), the agency stated that populations with diets rich
in these low saturated fat and low cholesterol foods experience many
health advantages, including lower rates of heart disease. In the
preamble to the proposed rule on dietary lipids and heart disease (56
FR 60727 at 60739), the agency stated that while total fat is not
directly linked to increased risk of CHD, it may have significant
indirect effects. Foods that are low in total fat facilitate reductions
in intakes of saturated fat and cholesterol to recommended levels.
Therefore, the agency tentatively concludes that proposed
Sec. 101.82(c)(2)(iii)(B) sets forth an appropriate requirement for
food to be eligible to bear the soy protein and CHD claim.
E. Optional Information
FDA is proposing in Sec. 101.82(d)(1) that the claim may state that
the development of heart disease depends on many factors and,
consistent with authorized CHD health claims, may list the risk factors
for heart disease that are listed in Secs. 101.75(d)(1), 101.77(d)(1),
and 101.81(d)(1). The agency is also proposing, in response to the
petition, that the claim may provide additional information about the
benefits of exercise and body weight management. This additional
information can provide a context that is useful for an understanding
of the relationship between soy protein and heart disease, but
manufacturers should be cautioned that it should not be presented in a
way that is misleading to the consumer.
In proposed Sec. 101.82(d)(2), consistent with Secs. 101.75(d)(2),
101.77(d)(2), and 101.81(d)(2), FDA is providing that the claim may
state that the relationship between a diet high in soy protein and
reduced risk of heart disease is through the intermediate link of
``blood cholesterol'' or ``blood total cholesterol'' and ``LDL-
cholesterol.'' The relationship between soy protein and reduced blood
total cholesterol and LDL-cholesterol is supported by the scientific
evidence presented in this proposal.
In Sec. 101.82(d)(3), the agency is proposing that, consistent with
Secs. 101.75(d)(3), 101.77(d)(3), and 101.81(d)(3), the claim may
include information from Sec. 101.82(a) and (b). These paragraphs
summarize information regarding the relationship between diets high in
soy protein and the risk of CHD and about the significance of that
relationship. This information helps to convey the seriousness of CHD
and the role that a diet high in soy protein can play to help reduce
the risk of CHD.
The agency is proposing that the claim may include any of the
optional information authorized to be included in Secs. 101.75(d)(5),
(d)(6), and (d)(7), 101.77(d)(5), (d)(6), and (d)(7), and 101.81(d)(5),
(d)(6), and (d)(7). The health claim may state that diets high in soy
protein and low in saturated fat and cholesterol are part of a dietary
pattern
[[Page 62993]]
that is consistent with dietary guidelines for Americans. The claim may
state that individuals with elevated serum lipids should consult their
physicians for medical advice and treatment and may include information
on the prevalence of CHD in the United States. The intent of this
information is to provide consumers with information that will help
them understand the seriousness of CHD in the United States and to help
them understand that diets high in soy protein are consistent with
dietary guidelines.
F. Model Health Claims
In proposed Sec. 101.82(e), FDA is providing model health claims to
illustrate the requirements of new Sec. 101.82. FDA emphasizes that
these model health claims are illustrative only. These model claims
illustrate the required, and some of the optional, elements of the
proposed rule. If the agency authorizes a claim about the relationship
between soy protein and CHD, manufacturers will be free to design their
own claim so long as it is consistent with Sec. 101.82(c).
In Secs. 101.82(e)(1) and (e)(2), the model claim illustrates all
of the required elements of the proposed health claim. The claim states
``25 grams of soy protein a day, as part of a diet low in saturated fat
and cholesterol, may reduce the risk of heart disease. A serving of
[name of food] supplies ---------- grams of soy protein.'' or ``Diets
low in saturated fat and cholesterol that include 25 grams of soy
protein may reduce the risk of heart disease. One serving of [name of
food] provides ---------- grams of soy protein.''
VI. Environmental Impact
The agency has determined under 21 CFR 25.30(k) that this action is
of a type that does not individually or cumulatively have a significant
effect on the human environment. Therefore, neither an environmental
assessment nor an environmental impact statement is required.
VII. Analysis of Impacts
A. Cost-Benefit Analysis
FDA has examined the impacts of the proposed rule under Executive
Order 12866. Executive Order 12866 directs Federal agencies to assess
all costs and benefits of available regulatory alternatives and, when
regulation is necessary, to select regulatory approaches that maximize
net benefits (including potential economic, environmental, public
health and safety, and other advantages; distributive impacts; and
equity). According to Executive Order 12866, a regulatory action is
``economically significant'' if it meets any one of a number of
specified conditions, including having an annual effect on the economy
of $100 million or adversely affecting in a material way a sector of
the economy, competition, or jobs. A regulation is considered
``significant'' under Executive Order 12866 if it raises novel legal or
policy issues. FDA finds that this proposed rule is neither an
economically significant nor a significant regulatory action as defined
by Executive Order 12866.
In addition, in accordance with the Small Business Regulatory
Enforcement Fairness Act (5 U.S.C. 801(a)(1)(A)(ii)), the Administrator
of the Office of Information and Regulatory Affairs of the Office and
Management and Budget (the Administrator) has determined that this
proposed rule is not a major rule for the purpose of congressional
review. A major rule for this purpose is defined in 5 U.S.C. 804(2) as
one that the Administrator has determined has resulted or is likely to
result in: (1) An annual effect on the economy of $100,000,000 or more;
or (2) a major increase in costs for prices for consumers, individual
industries, Federal, State, or local government agencies, or geographic
regions; or (3) significant adverse effects on competition, employment,
investment, productivity, innovation, or on the ability of U.S.-based
enterprises to compete with foreign-based enterprises in domestic or
export markets.
This proposed rule will give firms the option of making certain
label claims involving soy protein. No costs will be generated by this
proposed rule because it will not require any labels to be changed or
any product to be reformulated. Firms will only relabel or reformulate
products if the benefits to those firms outweigh the costs. Social
benefits may be generated by this proposed rule because the value some
consumers place on the information provided in these claims may be
greater than the cost to industry of making these claims. In general,
consumers may value this type of information because it will enable
them to eat a healthier diet. Consumers may value this type of
information presented on product labels, in particular, because it
would obviate the need to consult other sources of information and
because it may reassure consumers who are aware of the role of FDA in
regulating health claims on product labels that the information is
truthful, not misleading, and scientifically valid.
Consumer valuation of this information will reflect the value that
consumers place on reducing the likelihood of CHD and the perceived
usefulness of this information for reducing the likelihood of CHD.
However, consumers may either underestimate or overestimate the
usefulness of this information in reducing the likelihood of CHD.
Therefore, another metric for valuing the social benefits of this
proposed rule is the health care costs avoided by the reduction in CHD-
related disease and disability made possible by this proposed rule. If
consumers were aware of these health care costs and had an accurate
notion of the likelihood that such costs could be avoided by using the
information provided in the claims allowed by this proposed rule, then
consumer valuation of this information would be at least as great as
the value of any health care costs avoided. The value of the
information might be greater because some consumers might value the
information but might not choose to modify their behavior so as to
reduce the likelihood of CHD.
In general terms, the relevant regulatory options available to FDA
are as follows: (1) Allow this claim to be made under a broader set of
conditions than those specified in this proposed rule (e.g., with fewer
required elements in the claim, or with a lower level of soy protein in
a serving of food), and (2) allow this claim to be made under a more
restricted set of conditions than those specified in this rule (e.g.,
more required elements or higher levels of soy protein). Neither of
these alternatives would generate net costs because, like the proposed
action, firms would only relabel or reformulate products if the
benefits to those firms outweigh the costs. These options would
generate higher benefits than the proposed action if allowing this
claim to be made under either a broader set of conditions or more
restricted set of conditions than the proposed conditions would provide
consumers with more valuable information (that would nonetheless be
truthful, not misleading, and scientifically valid) or would make
possible a greater reduction in health care costs than would the
proposed action. FDA tentatively believes that no alternative
conditions exist that would render the net benefits of this proposed
rule greater than the proposed conditions. In particular, FDA believes
that the information proposed to be required in a health claim about
the relationship between soy protein and CHD is the minimum necessary
for the claim to be truthful, not misleading, and
[[Page 62994]]
scientifically valid, thereby maximizing the likelihood that qualifying
foods will be labeled with the claim and that consumers will be able to
use the information. Similarly, FDA believes that the amount of soy
protein proposed to be required for a food bearing this claim will
allow both the claim to appear on a significant number of foods and
consumers who use the claim, in the aggregate, to benefit from the use
of soy protein in their diet. However, FDA requests comments and
supporting information on any modifications of the conditions under
which this claim is allowed that would increase the net benefits of
this proposed rule.
B. Small Entity Analysis
FDA has examined the impacts of this proposed rule under the
Regulatory Flexibility Act (5 U.S.C. 601-612). The Regulatory
Flexibility Act requires Federal agencies to consider alternatives that
would minimize the economic impact of their regulations on small
businesses and other small entities. No costs will be generated by this
proposed rule because it will not require any labels to be changed, or
any product to be reformulated. Therefore, small businesses will only
relabel or reformulate products if the benefits (e.g., increased sales
of their products) to those small businesses outweigh the costs.
Accordingly, under the Regulatory Flexibility Act) (5 U.S.C. 605(b)),
the Commissioner of Food and Drugs certifies that this proposed rule,
if issued, will not have a significant economic impact on a substantial
number of small entities.
VIII. Paperwork Reduction Act
FDA tentatively concludes that the labeling requirements proposed
in this document are not subject to review by the Office of Management
and Budget because they do not constitute a ``collection of
information'' under the Paperwork Reduction Act of 1995 (44 U.S.C.
3501-3520). Rather, the proposed food labeling health claim on the
association between soy protein and coronary heart disease would be a
``public disclosure of information originally supplied by the Federal
government to the recipient for the purpose of disclosure to the
public'' (5 CFR 1320.3(c)(2)).
IX. Effective Date
FDA is proposing to make these regulations effective upon
publication in the Federal Register of a final rule based upon this
proposal.
X. Comments
Interested persons may, on or before January 25, 1999, submit to
the Dockets Management Branch (address above) written comments
regarding this proposal. Two copies of any comments are to be
submitted, except that individuals may submit one copy. Comments are to
be identified with the docket number found in brackets in the heading
of this document. Received comments may be seen in the office above
between 9 a.m. and 4 p.m., Monday through Friday.
XI. References
The following references have been placed on display in the Dockets
Management Branch (address above) and may be seen by interested persons
between 9 a.m. and 4 p.m., Monday through Friday.
1. Protein Technologies International, Inc., ``Health Claim
Petition,'' (CP1, vol. 1-3), May 4, 1998.
2. Protein Technologies International, Inc., ``Addendum to Health
Claim Petition,'' (CPI, vol. 4) August 10, 1998.
3. Pilch, S. M., Center for Food Safety and Applied Nutrition,
letter to M. M. Marcus, Protein Technologies International, Inc.,
August 12, 1998.
4. DHHS, Public Health Service (PHS), ``The Surgeon General's
Report on Nutrition and Health,'' U.S. Government Printing Office,
Washington, DC, pp. 83-137, 1988.
5. Food and Nutrition Board, National Academy of Sciences, ``Diet
and Health: Implications for Reducing Chronic Disease Risk,'' National
Academy Press, Washington, DC, pp. 291-309 and 529-547, 1989.
6. DHHS, Public Health Service (PHS) and National Institutes of
Health, ``National Cholesterol Education Program: Population Panel
Report,'' Bethesda, MD, pp. 1-40, 1990.
7. Sempos, C. T., J. I. Cleeman, M. D. Carroll, C. L. Johnson, P.S.
Bachorik, D. J. Gordon, V. L. Burt, R. R. Briefel, C. D. Brown, K.
Lippel, and B. M. Rifkind, ``Prevalence of High Blood Cholesterol Among
U.S. Adults. An Update Based on Guidelines From the Second Report of
the National Cholesterol Education Program Adult Treatment Panel,''
Journal of the American Medical Association, 269:3009-3014, 1993.
8. Ross, R., ``Atherosclerosis,'' in Cecil Textbook of Medicine,
edited by J. B. Wyndaarden, L. H. Smith, and J. C. Bennett, Harcourt
Brace Johanevich, Inc., Philadelphia, p. 293, 1992.
9. Van Sickle, G. J., G. K. Powell, P. J. McDonald, and R. M.
Goldblum, ``Milk- and Soy Protein-induced Enterocolitis: Evidence for
Lymphocyte Sensitization to Specific Food Proteins,'' Gastroenterology,
88:1915-1921, 1985.
10. Sampson, H. A. and S. M. Scanlon, ``Natural History of Food
Hypersensitivity in Children with Atopic Dermatitis,'' Journal of
Pediatrics, 115:23-27, 1989.
11. Roebuck, B. D., ``Trypsin Inhibitors: Potential Concern for
Humans?'' Journal of Nutrition, 117:398-400, 1987.
12. Anderson, R. L. and W. J. Wolf, ``Compositional Changes in
Trypsin Inhibitors, Phytic Acid, Saponins and Isoflavones Related to
Soybean Processing,'' Journal of Nutrition, 125:581S-588S, 1995.
13. Erdman, J. W., Jr., ``Effects of Soya Protein on Mineral
Availability,'' Journal of the American Oil Chemists' Society, 58:489-
493, 1981.
14. Liener, I., ``Significance for Humans of Biologically Active
Factors in Soybeans and Other Food Legumes,'' Journal of the American
Oil Chemists' Society, 56:121-129, 1979.
15. Kennedy, A. R., ``The Evidence for Soy Products as Cancer
Preventive Agents,'' Journal of Nutrition, 125:733S-743S, 1995.
16. Adlercreutz, H. and W. Mazur, ``Phyto-oestrogens and Western
Diseases,'' Annals of Medicine, 29:95-120, 1997.
17. Liener, I. E., ``Factors Affecting the Nutritional Quality of
Soya Products,'' Journal of the American Oil Chemists' Society, 58:406-
415, 1981.
18. Erdman, J. W., Jr. and E. J. Fordyce, ``Soy Products and the
Human Diet,'' American Journal of Clinical Nutrition, 49:725-737, 1989.
19. Breslau, N. A., L. Brinkley, K. D. Hill, and C. Y. C. Pak,
``Relationship of Animal Protein-rich Diet to Kidney Stone Formation
and Calcium Metabolism,'' Journal of Clinical Endocrinology and
Metabolism, 66:140-146, 1988.
20. Portale, A. A., B. P. Halloran, M. M. Murphy, and R. C. Morris,
Jr., ``Oral Intake of Phosphorus Can Determine the Serum Concentration
of 1,25-Dihydroxy-vitamin D by Determining its Production Rate in
Humans,'' Journal of Clinical Investigation, 77:46-55, 1986.
21. Moule, G. R., ``Clinical Aspects of Nutritional Infertility of
Ewes,'' Australian Veterinary Journal, 46(9):428-435, 1970.
22. Anthony, M. S., T. B. Clarkson, C. L. Hughes, Jr., T. M.
Morgan, and G. L. Burke, ``Soybean Isoflavones Improve Cardiovascular
Risk Factors Without Affecting the Reproductive System of Peripubertal
Rheusus Monkeys,'' Journal of Nutrition, 126:43-50, 1996.
23. Honore, E. K., J. K. Williams, M. S. Anthony, and T. B.
Clarkson, ``Soy
[[Page 62995]]
Isoflavones Enhance Coronary Vascular Reactivity in Atherosclerotic
Female Macaques,'' Fertility and Sterility, 67:148-154, 1997.
24. Cline, J. M., J. C. Paschold, M. S. Anathony, I. O. Obasanjo,
M. R. Adams, ``Effects of Hormonal Therapies and Dietary Soy
Phytoestrogens on Vaginal Cytology in Surgically Postmenopausal
Macaques,'' Fertility and Sterility, 65:1031-1035, 1996.
25. Lee, H. P., L. Gourley, S. W. Duffy, J. Esteve, J. Lee, and N.
E. Day, ``Dietary Effects on Breast-cancer Risk in Singapore,'' The
Lancet, 337:1197-1200, 1991.
26. Cassidy, A., S. Bingham, and K. R. Setchell, ``Biological
Effects of a Diet of Soy Protein Rich in Isoflavones on the Menstrual
Cycle of Premenopausal Women,'' American Journal of Clinical Nutrition,
60:333-340, 1994.
27. Bakhit, R. M., B. P. Klein, D. Essex-Sorlie, J. O. Ham, J. W.
Erdman, Jr., and S. M. Potter, ``Intake of 25 g of Soybean Protein With
or Without Soybean Fiber Alters Plasma Lipids in Men with Elevated
Cholesterol Concentrations,'' Journal of Nutrition, 124:213-222, 1994.
28. Baum, J. A., H. Teng, J. W. Erdman, Jr., R. M. Weigel, B. P.
Klein, V. W. Persky, S. Freels, P. Surya, R. M. Bakhit, E. Ramos, N. F.
Shay, and S. M. Potter, ``Long-term Intake of Soy Protein Improves
Blood Lipid Profiles and Increases Mononuclear Cell LDL Receptor mRNA
in Hypercholesterolemic Postmenopausal Women,'' American Journal of
Clinical Nutrition, in press.
29. Bosello, O., L. Cominacini, I. Zocca, U. Garbin, R. Compri, A.
Davoli, and L. Brunetti, ``Short- and Long-term Effects of Hypocaloric
Diets Containing Proteins of Different Sources on Plasma Lipids and
Apoproteins of Obese Subjects,'' Annals of Nutrition Metabolism,
32:206-214, 1988.
30. Carroll, K. K., P. M. Giovannetti, M. W. Huff, O. Moase, D. C.
K. Roberts, and B. M. Wolfe, ``Hypocholesterolemic Effect of
Substituting Soybean Protein for Animal Protein in the Diet of Healthy
Young Women,'' American Journal of Clinical Nutrition, 31:1312-1321,
1978.
31. Crouse, J. R. III, T. Morgan, J. G. Terry, J. Ellis, M.
Vitolins, and G. L. Burke, ``Effects of Soy Protein and Isoflavones on
Plasma Concentrations of Lipids and Lipoproteins,'' (submitted).
32. Cruz, M. L. A., W. W. Wong, F. Mimouni, D. L. Hachey, K. D. R.
Setchell, P. D. Klein, and R. C. Tsang, ``Effects of Infant Nutrition
on Cholesterol Synthesis Rates,'' Pediatric Research, 35:135-140, 1994.
33. Descovich, G. C., C. Ceredi, A. Gaddi, M. S. Benassi, G.
Mannino, L. Colombo, L. Cattin, G. Fontana, U. Senin, E. Mannarino, C.
Caruzzo, E. Bertelli, C. Fragiacomo, G. Noseda, M. Sirtori, and C. R.
Sirtori, ``Multicentre Study of Soybean Protein Diet for Outpatient
Hypercholesterolaemic Patients,'' Lancet, 2:709-712, 1980.
34. A. Gaddi, G. C. Descovich, G. Noseda, C. Fragiacomo, A.
Nicolini, G. Montanari, G. Vanetti, M. Sirtori, E. Gatti, and C. R.
Sirtori, ``Hypercholesterolaemia Treated by Soybean Protein Diet,''
Archives of Disease in Childhood, 62:274-278, 1987.
35. Gaddi, A., A. Ciarrocchi, A. Matteucci, S. Rimondi, G.
Ravaglia, G. C. Descovich, and C. R. Sirtori, ``Dietary Treatment for
Familial Hypercholesterolemia--Differential Effects of Dietary Soy
Protein According to the Apolipoprotein E Phenotypes,'' American
Journal of Clinical Nutrition, 53:1191-1196, 1991.
36. Giovannetti, P. M., K. K. Carroll, and B. M. Wolfe, ``Constancy
of Fasting Serum Cholesterol of Healthy Young Women Upon Substitution
of Soy Protein Isolate for Meat and Dairy Protein in Medium and Low Fat
Diets,'' Nutrition Research, 6:609-618, 1986.
37. Goldberg, A. P., A. Lim, J. B. Kolar, J. J. Grundhauser, F. H.
Steinke, and G. Schonfeld, ``Soybean Protein Independently Lowers
Plasma Cholesterol Levels in Primary Hypercholesterolemia,''
Atherosclerosis, 43:355-368, 1982.
38. Gooderham, M. J., H. Adlercreutz, S. T. Ojala, K. Wahala, and
B. J. Holub, ``A Soy Protein Isolate Rich in Genistein and Daidzein and
its Effects on Plasma Isoflavone Concentrations, Platelet Aggregation,
Blood Lipids and Fatty Acid Composition on Plasma Phospholipid in
Normal Men,'' Journal of Nutrition, 126:2000-2006, 1996.
39. Grundy, S. M. and J. J. Abrams, ``Comparison of Actions of Soy
Protein and Casein on Metabolism of Plasma Lipoproteins and Cholesterol
in Humans,'' American Journal of Clinical Nutrition, 38:245-252, 1983.
40. Holmes, W. L., G. B. Rubel, and S. S. Hood, ``Comparison of the
Effect of Dietary Meat Versus Dietary Soybean Protein on Plasma Lipids
of Hyperlipidemic Individuals,'' Atherosclerosis, 36:379-387, 1980.
41. Huff, M. W., P. M. Giovannetti, and B. M. Wolfe, ``Turnover of
Very Low-density Lipoprotein--Apoprotein B is Increased by Substitution
of Soybean Protein for Meat and Dairy Protein in the Diets of
Hypercholsterolemic Men,'' American Journal of Clinical Nutrition,
39:888-897, 1984.
42. Jacques, H., D. Laurin, S. Moorjani, F. H. Steinke, C. Gagne,
D. Brun, and P. J. Lupien, ``Influence of Diets Containing Cow's Milk
or Soy Protein Beverage on Plasma Lipids in Children with Familial
Hypercholesterolemia,'' Journal of the American College of Nutrition,
11(S):69S-73S, 1992.
43. Jenkins, D. J. A., T. M. S. Wolever, G. Spiller, G. Buckley, Y.
Lam, A. L. Jenkins, and R. G. Josse, ``Hypocholesterolemic Effect of
Vegetable Protein in a Hypocaloric Diet,'' Atherosclerosis, 78:99-107,
1989.
44. Kurowska, E. M., J. Jordan, J. D. Spence, S. Wetmore, L. A.
Piche, M. Radzikowski, P. Dandona, K. K. Carroll, ``Effects of
Substituting Dietary Soybean Protein and Oil for Milk Protein and Fat
in Subjects With Hypercholesterolemia,'' Clinical Investigative
Medicine, 20(3):162-170, 1997.
45. Laurin, D., H. Jacques, S. Moorjani, F. H. Steinke, C. Gagne,
D. Brun, and P. J. Lupien, ``Effects of a Soy-protein Beverage on
Plasma Lipoproteins in Children With Familial Hypercholesterolemia,''
American Journal of Clinical Nutrition, 54:98-103, 1991.
46. Lovati, M. R., C. Manzoni, A. Canavesi, M. Sirtori, V.
Vaccarino, M. Marchi, G. Gaddi, and C. R. Sirtori, ``Soybean Protein
Diet Increases Low Density Lipoprotein Receptor Activity in Mononuclear
Cells From Hypercholesterolemic Patients,'' Journal of Clinical
Investigation, 80:1498-1502, 1987.
47. Meinertz, H., O. Faergeman, K. Nilausen, M. J. Chapman, S.
Goldstein, and P. M. Laplaud, ``Effects of Soy Protein and Casein in
Low Cholesterol Diets on Plasma Lipoproteins in Normlipidemic
Subjects,'' Atherosclerosis, 72:63-70, 1988.
48. Meinertz, H., K. Nilausen, and O. Faergeman, ``Soy Protein and
Casein in Cholesterol-Enriched Diets: Effects on Plasma Lipoproteins in
Normolipidemic Subjects,'' American Journal of Clinical Nutrition,
50:786-793, 1989.
49. Mercer, N. J. H., K. K. Carroll, P. M. Giovannetti, F. H.
Steinke, and B. M. Wolfe, ``Effects on Human Plasma Lipids of
Substituting Soybean Protein Isolate for Milk Protein in the Diet,''
Nutrition Reports International, 35(2):279-287, 1987.
50. Meredith, L., M. Liebman, and K. Graves, ``Alterations in
Plasma Lipid Levels Resulting From Tofu and Cheese Consumption in Adult
Women,'' Journal of the American College of Nutrition, 8(6):573-579,
1989.
51. Potter, S. M., R. M. Bakhit, D. L. Essex-Sorlie, K. E.
Weingartner, K. M.
[[Page 62996]]
Chapman, R. A. Nelson, M. Prabhudesai, W. D. Savage, A. I. Nelson, L.
W. Winter, and J. W. Erdman, Jr., ``Depression of Plasma Cholesterol in
Men by Consumption of Baked Products Containing Soy Protein,'' American
Journal of Clinical Nutrition, 58:501-506, 1993.
52. Potter, J. M. and P. J. Nestel, ``Greater Bile Acid Excretion
With Soy Bean Than With Cow Milk In Infants,'' American Journal of
Clinical Nutrition, 29:546-551, 1976.
53. Sacks, F. M., J. L. Breslow, P. G. Wood, and E. H. Kass, ``Lack
of an Effect of Dairy Protein (Casein) and Soy Protein on Plasma
Cholesterol of Strict Vegetarians. An Experiment and a Critical
Review,'' Journal of Lipid Research, 24:1012-1020, 1983.
54. Shorey, R. L., B. Bazan, G. S. Lo, and F. H. Steinke,
``Determinants of Hypercholesterolemic Response to Soy and Animal
Protein-Based Diets,'' American Journal of Clinical Nutrition, 34:1769-
1778, 1981.
55. Sirtori, C. R., E. Agradi, F. Conti, O. Mantero, and E. Gatti,
``Soybean-protein Diet in the Treatment of Type-II
Hyperlipoproteinaemia,'' Lancet, 1:275-277, 1977.
56. Sirtori, C. R., C. Zucchi-Dentone, M. Sirtori, E. Gatti, G. C.
Descovich, A. Gaddi, L. Cattin, P. G. Da Col, U. Senin, E. Mannarino,
G. Avellone, L. Colombo, C. Fragiacomo, G. Noseda, and S. Lenzi,
``Cholesterol-lowering and HDL-raising Properties of Lecithinated Soy
Proteins in Type II Hyperlipidemic Patients,'' Annals of Nutrition
Metabolism, 29:348-357, 1985.
57. Steele, M. G., ``The Effect on Serum Cholesterol Levels of
Substituting Milk with a Soya Beverage,'' Australian Journal of
Nutrition and Dietetics, 49(1):24-28, 1992.
58. Van Raaij, J. M. A., M. B. Katan, J. G. A. J. Hautvast, and R.
J. J. Hermus, ``Effects of Casein Versus Soy Protein Diets on Serum
Cholesterol and Lipoproteins in Young Healthy Volunteers,'' American
Journal of Clinical Nutrition, 34:1261-1271, 1981.
59. Van Raaij, J. M. A., M. B. Katan, C. E. West, and J. G. A. J.
Hautvast, ``Influence of Diets Containing Casein, Soy Isolate, and Soy
Concentrate on Serum Cholesterol and Lipoproteins in Middle-aged
Volunteers,'' American Journal of Clinical Nutrition, 35:925-934, 1982.
60. Verrilo, A., A. de Teresa, P. C. Giarrusso, and S. La Rocca,
``Soybean Protein Diets in the Management of Type II
Hyperlipoproteinaemia,'' Atherosclerosis, 54:321-331, 1985.
61. Vessby, B., B. Karlstrom, H. Lithell, I. B. Gustafsson, and I.
Werner, ``The Effects on Lipid and Carbohydrate Metabolism of Replacing
Some Animal Protein by Soy-protein in a Lipid-Lowering Diet for
Hypercholesterolaemic Patients,'' Human Nutrition: Applied Nutrition,
36A:179-189, 1982.
62. Wang, M. F., S. Yamamoto, H. M. Chung, S. Y. Chung, S.
Miyatani, M. Mori, T. Okita, and M. Sugano, ``Antihypercholesterolemic
Effect of Undigested Fraction of Soybean Protein in Young Female
Volunteers,'' Journal of Nutrition Science Vitaminology, 41:187-195,
1995.
63. Widhalm, K., G. Brazda, B. Schneider, and S. Kohl, ``Effect of
Soy Protein Diet Versus Standard Low Fat, Low Cholesterol Diet on Lipid
and Lipoprotein Levels in Children With Familial or Polygenic
Hypercholesterolemia,'' Journal of Pediatrics, 123:30-34, 1993.
64. Wolfe, B. M., P. M. Giovannetti, D. C. H. Cheng, D. C. K.
Roberts, and K. K. Carroll, ``Hypolipidemic Effect of Substituting
Soybean Protein Isolate for all Meat and Dairy Protein in the Diets of
Hypercholesterolemic Men,'' Nutrition Reports International,
24:(6)1187-1198, 1981.
65. Nagata, C., N. Takatsuka, Y. Kurisu, and H. Shimizu,
``Decreased Serum Total Cholesterol Concentration is Associated With
High Intake of Soy Products in Japanese Men and Women,'' Journal of
Nutrition, 128:209-213, 1998.
66. Anderson, J.W., B.M. Johnstone, and M.E. Cook-Newell, ``Meta-
analysis of the Effects of Soy Protein Intake on Serum Lipids,'' New
England Journal of Medicine,, 333:276-282, 1995.
67. DHHS, National Heart, Lung, and Blood Institute, National
Cholesterol Education Porgram, ``Report of the Expert Panel on
Population Strategies for Blood Cholesterol Reduction (Population Panel
Report),'' Chapter II. Background and Introduction, NIH Publication No.
93-3046, Bethesda, MD, 1993.
68. Food and Nutrition Board, National Academy of Sciences, ``Diet
and Health: Implications for Reducing Chronic Disease Risk,'' National
Academy Press, Washington, DC, pp. 23-40, 1989.
69. U.S Preventive Services Task Force, ``Guide to Clinical
Preventive Services: An Assessment of the Effectiveness of 169
Interventions,'' Williams and Wilkins, Baltimore, MD, pp. xix-xxx,
1989.
70. Anthony, M. S., T. B. Clarkson, B. C. Bullock, and J. D.
Wagner, ``Soy Protein Versus Soy Phytoestrogens in the Prevention of
Diet-induced Coronary Artery Atherosclerosis of Male Cynomolgus
Monkeys,'' Arteriosclerosis, Thrombosis, and Vascular Biology,
17(11)2524-2531. 1997.
71. Nestle, P. J., T. Yamashita, T. Sashara, S. Pomeroy, A. Dart,
P. Komesaroff, A. Owen, and M. Abbey, ``Soy Isoflavones Improve
Systemic Arterial Compliance But Not Plasma Lipids in Menopausal and
Perimenopausal Women,'' Ateriosclerosis, Thrombosis, and Vascular
Biology, 17:3392-3398, 1997.
72. Sitori, C. R., E. Gianazza, C. Manzoni, M. R. Lovati, and P. A.
Murphy, ``Role of Isoflavones in the Cholesterol Reduction of Soy
Proteins in the Clinic,'' letter to the editor, American Journal of
Clinical Nutrition, 65:166-167, 1997.
List of Subjects in 21 CFR Part 101
Food labeling, Incorporation by reference, Nutrition, Reporting and
recordkeeping requirements.
Therefore, under the Federal Food, Drug, and Cosmetic Act and under
authority delegated to the Commissioner of Food and Drugs, it is
proposed that 21 CFR part 101 be amended as follows:
PART 101--FOOD LABELING
1. The authority citation for 21 CFR part 101 continues to read as
follows:
Authority: 15 U.S.C. 1453, 1454, 1455; 21 U.S.C. 321, 331, 342,
343, 348, 371.
2. New Sec. 101.82 is added to subpart E to read as follows:
Sec. 101.82 Health claims: Soy protein and risk of coronary heart
disease (CHD).
(a) Relationship between diets that are low in saturated fat and
cholesterol and that include soy protein and the risk of CHD. (1)
Cardiovascular disease means diseases of the heart and circulatory
system. CHD is one of the most common and serious forms of
cardiovascular disease and refers to diseases of the heart muscle and
supporting blood vessels. High blood total cholesterol and low density
lipoprotein (LDL)-cholesterol levels are associated with increased risk
of developing CHD. High CHD rates occur among people with high total
cholesterol levels of 240 milligrams per deciliter (mg/dL) (6.21
(millimole per liter (mmol/L))) or above and LDL-cholesterol levels of
160 mg/dL (4.13 mmol/L) or above. Borderline high risk total
cholesterol levels range from 200 to 239 mg/dL (5.17 to 6.18 mmol/L)
and 130 to 159 mg/dL (3.36 to 4.11 mmol/L) of LDL-cholesterol. The
scientific evidence establishes that diets high in saturated fat and
cholesterol are associated with increased levels of blood total and
LDL-cholesterol and, thus, with increased risk of CHD.
[[Page 62997]]
(2) Populations with a low incidence of CHD tend to have
relatively low blood total cholesterol and LDL-cholesterol levels.
These populations also tend to have dietary patterns that are not only
low in total fat, especially saturated fat and cholesterol, but are
also relatively high in plant foods that contain dietary fiber and
other components.
(3) Scientific evidence demonstrates that diets low in saturated
fat and cholesterol may reduce the risk of CHD. Other evidence
demonstrates that the addition of soy protein to a diet that is low in
saturated fat and cholesterol may also help to reduce the risk of CHD.
(b) Significance of the relationship between diets that are low in
saturated fat and cholesterol and that include soy protein and the risk
of CHD. (1) CHD is a major public health concern in the United States.
It accounts for more deaths than any other disease or group of
diseases. Early management of risk factors for CHD is a major public
health goal that can assist in reducing risk of CHD. High blood total
and LDL-cholesterol are major modifiable risk factors in the
development of CHD.
(2) Intakes of saturated fat exceed recommended levels in the
diets of many people in the United States. One of the major public
health recommendations relative to CHD risk is to consume less than 10
percent of calories from saturated fat and an average of 30 percent or
less of total calories from all fat. Recommended daily cholesterol
intakes are 300 mg or less per day. Scientific evidence demonstrates
that diets low in saturated fat and cholesterol are associated with
lower blood total and LDL-cholesterol levels. Soy protein, when
included in a low saturated fat and cholesterol diet, also helps to
lower blood total and LDL-cholesterol levels.
(c) Requirements. (1) All requirements set forth in Sec. 101.14
shall be met.
(2) Specific requirements--(i) Nature of the claim. A health claim
associating diets that are low in saturated fat and cholesterol and
that include soy protein with reduced risk of heart disease may be made
on the label or labeling of a food described in paragraph (c)(2)(iii)
of this section, provided that:
(A) The claim states that diets that are low in saturated fat and
cholesterol and that include soy protein ``may'' or ``might'' reduce
the risk of heart disease;
(B) In specifying the disease, the claim uses the following terms:
``heart disease'' or ``coronary heart disease'';
(C) In specifying the substance, the claim uses the term ``soy
protein'';
(D) In specifying the fat component, the claim uses the terms
``saturated fat'' and ``cholesterol'';
(E) The claim does not attribute any degree of risk reduction for
CHD to diets that are low in saturated fat and cholesterol and that
include soy protein;
(F) The claim does not imply that consumption of diets that are low
in saturated fat and cholesterol and that include soy protein is the
only recognized means of achieving a reduced risk of CHD; and
(G) The claim specifies the daily dietary intake of soy protein
that is necessary to reduce the risk of coronary heart disease and the
contribution one serving of the product makes to the specified daily
dietary intake level. The daily dietary intake level of soy protein
that has been associated with reduced risk of coronary heart disease is
25 grams (g) or more per day of soy protein.
(ii) Nature of the substance. (A) Soy protein from the legume seed
Glycine max.
(B) FDA will measure soy protein by method No. 988.10 from the
``Official Methods of Analysis of the Association of Official
Analytical Chemists International,'' 16th Ed. (1995), which is
incorporated by reference in accordance with 5 U.S.C. 522(a) and 1 CFR
part 51. Copies may be obtained from the Association of Official
Analytical Chemists International, 481 North Frederick Ave., suite 500,
Gaithersburg, MD 20877-2504, or may be examined at the Center for Food
Safety and Applied Nutrition's Library, 200 C St. SW., rm. 3321,
Washington, DC, or at the Office of the Federal Register, 800 North
Capitol St. NW., suite 700, Washington, DC;
(iii) Nature of the Food Eligible to Bear the Claim. (A) The food
product shall contain at least 6.25 g of soy protein reference amount
customarily consumed of the food product;
(B) The food shall meet the nutrient content requirements in
Sec. 101.62 for a ``low saturated fat,'' ``low cholesterol,'' and ``low
fat'' food.
(d) Optional information. (1) The claim may state that the
development of heart disease depends on many factors and may identify
one or more of the following risk factors for heart disease about which
there is general scientific agreement: A family history of CHD;
elevated blood total and LDL-cholesterol; excess body weight; high
blood pressure; cigarette smoking; diabetes; and physical inactivity.
The claim may also provide additional information about the benefits of
exercise and management of body weight to help lower the risk of heart
disease;
(2) The claim may state that the relationship between intake of
diets that are low in saturated fat and cholesterol and that include
soy protein and reduced risk of heart disease is through the
intermediate link of ``blood cholesterol'' or ``blood total and LDL-
cholesterol;''
(3) The claim may include information from paragraphs (a) and (b)
of this section, which summarize the relationship between diets that
are low in saturated fat and cholesterol and that include soy protein
and CHD and the significance of the relationship;
(4) The claim may state that a diet low in saturated fat and
cholesterol that includes soy protein is consistent with ``Nutrition
and Your Health: Dietary Guidelines for Americans,'' U.S. Department of
Agriculture (USDA) and Department of Health and Human Services (DHHS),
Government Printing Office (GPO);
(5) The claim may state that individuals with elevated blood total
and LDL-cholesterol should consult their physicians for medical advice
and treatment. If the claim defines high or normal blood total and LDL-
cholesterol levels, then the claim shall state that individuals with
high blood cholesterol should consult their physicians for medical
advice and treatment;
(6) The claim may include information on the number of people in
the United States who have heart disease. The sources of this
information shall be identified, and it shall be current information
from the National Center for Health Statistics, the National Institutes
of Health, or ``Nutrition and Your Health: Dietary Guidelines for
Americans,'' USDA and DHHS, GPO;
(e) Model health claim. The following model health claims may be
used in food labeling to describe the relationship between diets that
are low in saturated fat and cholesterol and that include soy protein
and reduced risk of heart disease:
(1) 25 grams of soy protein a day, as part of a diet low in
saturated fat and cholesterol, may reduce the risk of heart disease. A
serving of [name of food] supplies -------------- grams of soy protein.
(2) Diets low in saturated fat and cholesterol that include 25
grams of soy protein may reduce the risk of heart disease. One serving
of [name of food] provides ---------------- grams of soy protein.
Dated: November 2, 1998.
William B. Schultz,
Deputy Commissioner for Policy.
[FR Doc. 98-30008 Filed 11-9-98; 8:45 am]
BILLING CODE 4160-01-F
[[Page 62998]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.001
[[Page 62999]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.002
[[Page 63000]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.003
[[Page 63001]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.004
[[Page 63002]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.005
[[Page 63003]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.006
[[Page 63004]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.007
[[Page 63005]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.008
[[Page 63006]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.009
[[Page 63007]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.010
[[Page 63008]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.011
[[Page 63009]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.012
[[Page 63010]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.013
[[Page 63011]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.014
[[Page 63012]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.015
[[Page 63013]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.016
[[Page 63014]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.017
[[Page 63015]]
[GRAPHIC] [TIFF OMITTED] TP10NO98.018
[FR Doc. 98-30008 Filed 11-9-98; 8:45 am]
BILLING CODE 4160-01-C