Federal Register / Vol. 58, No. 3 / Wednesday, January 6, 1993
   / Rules and Regulations
   DEPARTMENT OF HEALTH AND HUMAN SERVICES
   Food and Drug Administration
   
   
   21 CFR Part 101
   
   [Docket No. 91N - 0099]
   RIN 0905 - AB67
   
   
   Food Labeling: Health Claims and Label Statements; Dietary
   Fiber and Cardiovascular Disease
   
   AGENCY: Food and Drug Administration, HHS.
   
   ACTION: Final rule.
   
                                                                  
                                                                  
                                                                  
                                                       
   
   
   SUMMARY: The Food and Drug Administration (FDA) is announcing
   its decision not to authorize the use on the label or labeling
   of foods of health claims relating to an association between
   dietary fiber and cardiovascular disease (CVD). However, FDA
   is authorizing 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 coronary heart disease (CHD).
   This action is in response to provisions of the Nutrition
   Labeling and Education Act of 1990 (the 1990 amendments) that
   bear on health claims, and was developed in accordance with
   the final rule on general requirements for health claims,
   published elsewhere in this issue of the Federal Register.
       On the basis of the totality of the publicly available
   scientific evidence, including recently available evidence,
   the agency has concluded that there is significant scientific
   agreement among qualified experts that a claim relating diets
   low in saturated fat and cholesterol, and high in fruits,
   vegetables, and grain products that contain soluble fiber, to
   reduced risk of CHD is supported. The evidence is not
   sufficient to attribute the reduction in risk to soluble fiber
   or to a specific type or characteristic of soluble fiber, or
   to other components of these diets.
   
   EFFECTIVE DATE: May 8, 1993.
   
   FOR FURTHER INFORMATION CONTACT: Joyce J. Saltsman, Center for
   Food Safety and Applied Nutrition (HFS - 165), Food and Drug
   Administration, 200 C St. SW., Washington, DC 20204, 202 - 205
   - 5916.
   
   SUPPLEMENTARY INFORMATION:
   
   I. Background
   
       In the Federal Register of November 27, 1991 (56 FR
   60582), FDA proposed to deny the use on food labeling of
   health claims relating diets high in dietary fiber to reduced
   risk of CVD. The proposed rule was issued under provisions of
   the 1990 amendments (Pub. L. 101 - 535) that bear on health
   claims and in accordance with the proposed general
   requirements for health claims for food (56 FR 60537, November
   27, 1991). As amended in 1990, the Federal Food, Drug, and
   Cosmetic Act (the act) providesthat a food is misbranded if it
   bears a claim that characterizes the relationship of a
   nutrient to a disease or health-related condition unless the
   claim is made in accordance with sections 403(r)(3) or
   (r)(5)(D) of the act (21 U.S.C. 343(r)(3) or (r)(5)(D)).
       Section 3(b)(1)(A) of the 1990 amendments specifically
   requires that the agency determine whether health claims
   respecting 10 nutrient/disease relationships meet the
   requirements of section 403(r)(3) or (r)(5)(D) of the act. The
   relationship betweendietary fiber and CVD is one of the claims
   required to be evaluated. In the proposed rule published in
   the Federal Register of November 21, 1991 (56 FR 60582), FDA
   limited its review of the science to those aspects of the
   dietary fiber/CVD relationship for which the strongest
   scientific evidence and agreement already existed: dietary
   soluble fiber and CHD.
       In the proposed rule, FDA requested written comments on
   its tentative determination not to authorize a health claim
   for dietary fiber and CVD. FDA also requested comments on the
   following issues: (1) Should the agency permit a claim on the
   label or in labeling of foods that states that diets high in
   fruit, vegetables, and whole grains are associated with a
   reduced risk of certain forms of cancer and CVD?; (2) If such
   statements should be permitted, what criteria should be used
   to identify foods thatfor foods as well as nutrients?
       On January 30 and 31, 1992, FDA held public hearings on
   all aspects of the proposed rules published in response to the
   1990 amendments, including health claims for dietary fiber and
   CVD. In addition, because of new evidence identified from
   literature searches and new data submitted with comments to
   the proposed rule, FDA reopened the comment period on dietary
   fiber and CVD for 30 days (57 FR 32751, July 23, 1992).
       In response to its proposed rule on dietary fiber and CVD,
   the agency received approximately 130 comments from consumers,
   consumer advocacy groups, State health departments,
   organizations of health professionals, the food industry, and
   Government agencies. A number of comments were received that
   were more appropriately answered in other dockets, and these
   were forwarded to the appropriate docket for response.
       Most of the comments specific to the proposal for a health
   claim for fiber and CVD provided explanations in support of or
   in opposition to provisions of the proposed regulation. Some
   of the comments contained relevant scientific studies not
   includedin the agency's proposed rule. These additional
   studies and those identified through literature searches that
   had not been previously reviewed in the proposal (56 FR 60582)
   are included in the agency's review below. The agency has
   summarized and responded to issues raised in the comments in
   section III. of this document.
   
   II. Updated Review of Scientific Evidence
   
       FDA, to ensure that it had not overlooked new and
   significant scientific data, reviewed human studies published
   since the publication of the proposed rule that it had
   identified in a standard literature search. In addition, FDA
   carefully reviewed all relevant scientific data submitted as
   comments. The availability of the new data was announced in a
   notice of reopening of comment period (57 FR 32751).
   
   A. Human Studies
   
       Detailed summaries of studies discussed below are
   presented in Table 1. FDA's explanation for separately
   evaluating studies on mildly to moderately
   hypercholesterolemic individuals and normocholesterolemic
   individuals is found in its proposed rule (56 FR 60587).
   Multiple sources of soluble fiber, including oat bran and
   other cereal brans, legumes, pectin, psyllium, and guar gum,
   were used in these studies.
   1. Hypercholesterolemics: ``typical'' or ``usual'' diets
       Leadbetter et al. (Ref. 83) evaluated the
   hypocholesterolemic effects of increasing intakes of [-glucan,
   the major type of dietary soluble fiber in oat bran. A
   four-by-four Latin square design was used in this randomized
   intervention study with40 hypercholesterolemic (total serum
   cholesterol 250 to 348 milligram/deciliter (mg/dL)) men and
   women, ages 25 to 64, in New Zealand. Subjects added 0, 30,
   60, and 90 grams/day (g/day) oat bran to their usual diet for
   1-month intervals. There was no wash-out between periods. Oat
   bran was provided in weighed packages and detailed advice and
   recipes were provided on how to incorporate it into the diet.
   The total dietary fiber content of the regular diets, without
   oat bran, ranged from 23 to 27 g/day. Results showed no
   significant effect of oat bran on serum cholesterol at any
   dose. There was no dose-related trend and no correlation
   between bran dose and changes in serum cholesterol. The
   authors stated that the oat bran used in this study was lower
   in soluble fiber (3.7 to 4.2 percent [-glucan) than oat bran
   used in studies showing a significant lowering of serum
   cholesterol with oat bran supplementation.
       Cara et al. (Ref. 84) evaluated the hypocholesterolemic
   properties of wheat germ in 10 hypercholesterolemic men and
   women (serum cholesterol 254 to 367 mg/dL), ages 35 to 68
   years. Subjects consumed their regular diets for 1 week, then
   added 30 g/daysed significantly (8.6 percent) after wheat germ
   intervention and returned to baseline during the 1 month
   followup period. Dietary soluble fiber and total saturated fat
   before and during the treatment period were not reported. The
   authors speculate thatthe high vegetable protein content of
   wheat germ could account for the observed results.
       Karlander et al. (Ref. 85) evaluated the
   hypocholesterolemic properties of beet fiber in 13
   hypercholesterolemic noninsulin-dependent diabetics mellitus
   (NIDDM) men and women (mean serum cholesterol of 275 mg/dL).
   Five subjects were on chronic beta blockers and diuretics and
   eight were on diet treatment with sulfonylurea (SU). This was
   a controlled, randomized intervention trial with cross-over
   design. The study was divided into three 6-week periods with a
   run-in period followed by either fiber intervention (20 g/day
   beet fiber) or the subject's regular diet for 6 weeks, then
   cross-over to the other diet. Obese subjects were given
   dietary advice to aid in weight control. Results showed no
   significant difference in total serum cholesterol between
   control and fiber periods for subjects advised to reduce
   energy intake. The SU group showed significantly decreased (10
   percent) serum cholesterol during the fiber phase (total
   cholesterol decreased from 275 to 247 mg/dL). The SU group had
   a slight but significant loss of body mass during the run-in
   period only. There was no effect on serum triglycerides.
       Spiller et al. (Ref. 87) evaluated the
   cholesterol-lowering properties of guar gum compared to oat
   bran in a 3-week, intervention trial with cross-over. Thirteen
   men and women, mean age 62 years, with mild to moderate
   hypercholesterolemia (serum cholesterol 204 to 276 mg/dL),
   were randomized to receive either 15 g (11 g dietary fiber, of
   which there was 10 g soluble fiber) of guar gum per day or 77
   g/day (11 g dietary fiber, 5 g soluble fiber) oat fiber
   source, divided into three servings. The fibers were provided
   in weighed packets with instructions to mix the fiber with
   water or juice and consume before each main meal. After 21
   days, subjects switched to the other fiber source. There was
   no wash-out between test periods. Blood samples were collected
   on days 14 and 21 during treatment periods. Results showed a
   significant reduction in serum cholesterol, compared to
   baseline, for both groups. Guar reduced serum cholesterol 11
   percent and oat fiber 6 percent. Maximum cholesterol reduction
   was experienced after 14 days on the test fiber. No
   significant change occurred in serum cholesterol between days
   14 and 21 on either test fiber, with serum cholesterol values
   increasing slightly but not significantly. Factors confounding
   these results include small sample size, the absence of a
   wash-out between test periods, and short treatment periods.
   Dietary intakes of soluble fiber were not reported.
       Kawatra et al. (Ref. 88) reported significant reductions
   in total serum cholesterol and low density lipoprotein
   (LDL)-cholesterol in 20 overweight Indonesian men and women
   with mild to moderate hypercholesterolemia consuming guar gum.
   Subjects consumed 15 g/day guar gum with their normal diet for
   6 weeks. The guar was consumed 15 minutes before the main meal
   in the form of biscuits (10 g guar) and mixed (5 g guar gum)
   with a flavored drink. Total serum cholesterol decreased 16.7
   percent and LDL-cholesterol decreased 26.5 percent. Intake of
   total dietary fiber, dietary soluble fiber, and saturated fat
   were not reported.
       Tinker et al. (Ref. 89) evaluated the cholesterol-lowering
   properties of prunes in a randomized, cross-over trial.
   Forty-one men, ages 29 to 79, with mild to moderate
   hypercholesterolemia (serum cholesterol 201 to 290 mg/dL)
   consumed either grape juice (360 milliliter (mL)/day) or 12
   prunes (100 g/day) for 4 weeks followed by cross-over to other
   diet for an additional 4 weeks. Prunes provided 6 to 7 g of
   total dietary fiber and 3.6 to 4.2 g of soluble fiber as
   pectin. Base diets included 18 g of totae prune period. Four
   to five percent of the energy was from alcohol. There was no
   significant difference between serum cholesterol on the prune
   diet and on the grape juice diet.
       A final report by Earll et al., July 1986 (Ref. 90), was
   submitted with one of the comments. This was an intervention
   study with seven free living hyperlipidemic (serum cholesterol
   ranged from 261 to 346 mg/dL) male and female patients.
   Subjects were asked to consume 24 g/day of corn fiber
   (containing less than 2 g soluble fiber) for 6 weeks, then 48
   g/day corn fiber for an additional 6 weeks with their regular
   diet. The test period was followed by an 8-week wash-out
   period. Actual consumption of corn fiber was slightly less
   according to records kept by the subjects. Total cholesterol
   was reduced, on average from 298 to 253 mg/dL (significant),
   although much smaller changes were seen in two subjects, and
   one subject had an increase in serum cholesterol during the
   study period. In all but one of the subjects, serum
   cholesterol remained above 200 mg/dL with fiber intervention,
   suggesting dietary therapy was not adequate. The dietary
   intakes of the subjects before and during the test period were
   notrecorded.
       Whyte et al. (Ref. 104) reported significant reductions in
   serum cholesterol in 23 men with mild hypercholesterolemia
   (total cholesterol ranged from 209 to 259 mg/dL) consuming oat
   bran. Subjects consumed 123 g/day oat bran or 54 g/day wheat
   bran cereal with their regular diets for 4 weeks followed by
   cross-over to the other fiber cereal. All subjects consumed
   wheat cereal during a 3-week baseline period prior to
   randomization to test groups. Total dietary fiber and fat
   intake were approximately the same between groups. Total serum
   cholesterol and LDL-cholesterol, when consuming oat bran,
   decreased 4 percent and 5.5 percent, respectively, compared to
   wheat bran. The authors note that one of their earlier studies
   with 12 g of oat bran showed significant decreases in serum
   cholesterol of 6 percent compared to 12 g of wheat bran.
   According to their analyses, both oat brans were similar in
   composition (Ref. 104). The smaller decrease in serum
   cholesterol reported in this study as compared to some other
   oat bran studies, not conducted by these authors, may be due
   to the difference in [-glucan content of different strains of
   oats. This 4-week study does not address long-term
   effectiveness of oat bran intervention.
       Uusitupa et al. (Ref. 100), in a randomly-allocated,
   double-blind parallel group trial, tested the effects of guar
   gum on 39 individuals with NIDDM (mean serum cholesterol of
   groups: 253 and 237 mg/dL). The test group received 5 g of
   guar gum three times per day (estimated total of 10 g soluble
   fiber) before meals for 3 months, while the control group
   received 5 g of wheat flour 3 times per day before meals.
   After 3 months, the control group was switched to guar and
   both groups were followed for anadditional 10 months. At the
   end of 3 months, the guar group showed a significant lowering
   of serum total cholesterol. Over the remaining 10 months, the
   group which began as the test group had an increase in total
   cholesterol, although it remained significantly lower than
   prior to the trial. The group which began as the control
   group, but switched to guar, demonstrated the lowest serum
   cholesterol during month 5 (208 mg/dL). This was followed by
   increasing serum cholesterol to a maximum at month 11 (242
   mg/dL). After 12 months, serum cholesterol for this group was
   233 mg/dL. Significant weight loss occurred in both the
   control and treatment groups to a similar degree. The dietary
   intakes for both groups were not reported.
       Spencer and Gee (Ref. 109) evaluated the
   cholesterol-lowering properties of apple juice supplemented
   with 10 g of dietary fiber (70 percent soluble fiber,
   predominantly from gum arabic) versus plain apple juice.
   Thirty-one mildly hypercholesterolemic there was a significant
   decrease in serum total cholesterol and LDL-cholesterol during
   the period of consumption of fiber-enriched juice. When order
   of presentation is considered, there were inconsistencies in
   the changes. The group which began the trial with the placebo
   did not show any change from baseline in serum cholesterol
   during the placebo phase, whereas the group which received the
   placebo during the second half of the trial maintained the
   lower cholesterol level which occurred during the fiber
   supplementation period. Cholesterol intake was significantly
   higher in the juice-only group.
       Niemi et al. (Ref. 99), in a double-blind, cross-over
   trial, reported significant lowering of serum cholesterol in a
   group that received 15 g/day guar gum (estimated total of 10 g
   soluble fiber) as a supplement for 12 weeks when compared to
   the group which received cellulose, the placebo. The 16 women
   and 6 men chosen as subjects were all (NIDDM) between the ages
   of 40 and 76. Nineteen of these subjects were on medication
   for their diabetes. Although the patients were advised to
   maintain their normal diets, no dietary measurements were
   reported to verify this.
       Kirsten et al. (Ref. 97) evaluated the hypocholesterolemic
   properties of guar gum in 13 men and women with type IIa and
   IIb hyperlipidemia (serum cholesterol concentration  251
   mg/dL). The study was divided into three phases: the 30-day
   pretreatment phase (baseline), the 60-day treatment, and the
   60-day post-treatment period. During the treatment phase,
   subjects consumed 4 g of guar gum, mixed with water or juice
   before breakfast, lunch, and dinner (12 g guar/day,
   estimated--8 g soluble fiber). Subjects were only instructed
   to avoid cholesterol-rich foods during the study period. Both
   total serum cholesterol and LDL-cholesterol decreased
   significantly compared to pretreatment levels. During the
   post-treatment period, serum cholesterol returned tobaseline
   and LDL-cholesterol increased above baseline. The dietary
   intakes of the subjects during each phase of the study were
   not reported.
       Cerda et al. (Ref. 105) evaluated the hypocholesterolemic
   effects of grapefruit pectin in a double-blind,
   placebo-controlled, cross-over study. Twenty-seven
   hypercholesterolemic men and women (mean serum cholesterol of
   275 mg/dL) consumed 15 g of grapefruit pectin (27 tablets, 9
   per meal) or 15 g of flour (also in tablet form) per day for 4
   weeks followed by a 4-week wash-out before cross-over. During
   the pectin period total cholesterol decreased by almost 8
   percent and LDL-cholesterol by 11 percent (both significantly
   lower compared to baseline). There was no change in serum
   cholesterol during the placebo period. The dietary intake of
   the subjects during each period of the study was not reported.
   The short test periods do not address long-term usefulness of
   pectin.
       Haskell et al. (Ref. 106) evaluated the
   hypocholesterolemic properties of four isolated, purified
   soluble fibers in four separate trials. All subjects had serum
   cholesterol levels between 200 and 280 mg/dL and were
   randomized to 1 of the four studies. Study one was a 12-week
   intervention trial using a powdered soluble fiber mixture
   providing 17.2 g/day of soluble fiber. The powder consisted of
   acacia gum (9.7 g), psyllium (4.9 g), and guar gum (2.6 g).
   The soluble fiber mixture was tested against afructose placebo
   (15 g/day). Results showed no statistically significant
   changes from baseline to 6 or 12 weeks within or between
   groups. Study two was a 4-week intervention testing 15 g/day
   of acacia gum powder against 15 g/day fructose powder. There
   was no significant change from baseline or compared to the
   placebo. Study three was an 8-week cross-over trial using 10
   g/day of guar (estimated 6.7 g soluble fiber) as a control and
   a 15 g/day fiber mixture of pectin (3.9 g), psyllium (6.3 g),
   guar (3.3 g)ximately 14 percent and 12 percent, respectively),
   and high density lipoprotein (HDL)-cholesterol (about 6
   percent) significantly. Study four was a 4-week dose-response
   study with increasing amounts of soluble fiber from a mixture
   of pectin, psyllium, guar, and locust bean gum. Three groups
   received either 5 g, 10 g, or 15 g/day of the mixture. Results
   showed that the serum cholesterol and LDL-cholesterol of the
   group consuming 15 g/day was significantly lower than the
   placebo.
       A 1984 study by Anderson et al. (Ref. 110) was submitted
   with comments in support of a long-term hypocholesterolemic
   effect of soluble fiber from oat bran and beans. Ten men, ages
   46 to 66, with serum cholesterol above 260 mg/dL were randomly
   assigned to oat bran or bean-supplemented diets for 21 days
   following a 7-day baseline period on a metabolic ward. The
   base diet was a ``typical'' American diet with 38 percent of
   calories as fat and 450 mg cholesterol. To this diet, the
   subjects added either 100 g of oat bran or 100 g of dried
   beans per day. Both of these diets provided 18 g of total
   soluble fiber and 48 to 50 g of total plant fiber per day.
   Before discharge, the subjects were instructed on the use of a
   high fiber (50 g of total fiber per day) maintenance diet with
   oat or bean product supplements at home. The high fiber diet
   was also low in fat (30 percent of calories as fat), low in
   saturated fat (10 percent of calories), and low in cholesterol
   (150 mg/day). Ten men were followed on their home diets for 24
   weeks and 4 for 99 weeks. Results of the metabolic ward phase
   of the study showed that the oat bran and bean diets lowered
   serum cholesterol significantly (23 percent) over the 3-week
   period compared to baseline. At 24 weeks (the followup phase),
   serum cholesterol levels were significantly (26 percent) lower
   than baseline. Cholesterol levels in the four men followed for
   99 weeks were 23 percent lower than baseline (significant at
   p<0.0025). Reductions in LDL-cholesterol were also significant
   during both phases of the study. HDL-cholesterol decreased
   significantly (20 percent) during the metabolic ward phase but
   increased during the long-term followup. All 10 subjects lost
   approximately 4 pounds (lb) (significant at p<0.0025) during
   the metabolic ward phase of the study. The investigators
   reported that changes in body weight were not significantly
   correlated with the changes in serum cholesterol. An
   additional 4 lb of weight was lost during the 24-week phase of
   followup.
       An unpublished study (Ref. 119) evaluated the cholesterol
   lowering properties of oat gum in hypercholesterolemics (mean
   serum cholesterol of 255 mg/dL). Instant oat gum (3.6 g) or a
   placebo (maltodextrin) were mixed with a noncarbonated diet
   fruit drink (250 mL) and consumed twice a day at each main
   meal for 4 weeks. There was a 3-week wash-out between
   treatment periods and after the last oat gum period. Subjects
   were randomly assigned to start the treatment period with
   either the oat gum beverage or the placebo. Results showed
   significantly lower serum cholesterol after 4 weeks on oat gum
   compared to both the baseline (p = 0.02) and the placebo (p =
   0.001). Although subjects were asked to maintain their weight,
   subjects' weights were not reported.
       Bridges et al. (Ref. 120) evaluated the effect of oat bran
   on serum cholesterol and serum acetate in hypercholesterolemic
   men admitted to a metabolic ward. Animal studies have shown
   that both acetate and propionate inhibit cholesterol synthesis
   (Ref. 120). The 20 subjects were divided into two groups:
   Wheat bran group (mean serum cholesterol of 252 mg/dL) and oat
   bran group (mean serum cholesterol of 305 mg/dL). Following 1
   week on a typical American diet, the diets were supplemented
   for 21 days with either 110 g of oat bran per day or 40 g of
   wheat bran. Results showed that the oat bran group experienced
   significantly lower (p = 0.05) serum cholesterol than the
   wheat bran group. However, in the wheat bran group, the
   baseline cholesterol level had bt (p = 0.001) weight loss in
   both groups. The weight loss appeared to be greater in the oat
   bran group. LDL-cholesterol was significantly lower (p =
   0.005) in the oat bran group as compared to the group's
   pretreatment values. There was no significant difference in
   LDL-cholesterol between groups. Serum acetate values were
   significantly higher in the oat bran group than in the
   pretreatment diets. Wheat bran did not change serum acetate
   significantly compared to the pretreatment diets.
       Kashtan et al. (Ref. 121) evaluated the effects of wheat
   bran and oat bran supplements on blood lipids and lipoproteins
   in 84 subjects with mild hypercholesterolemia. This was a
   controlled, parallel, double-blind study in which subjects
   consumed either oat bran supplements (with 11 to 17 g dietary
   fiber and an estimated 5 to 8 g of soluble fiber) or a wheat
   bran cream of wheat mixture (11 to 17 g of dietary fiber) each
   day for 14 days. Defined diets were delivered to the subjects'
   homes. The diets provided one of four energy amounts: 1,600,
   2,000, 2,400, and 2,800 calories, with 37 percent of energy as
   fat, 47 percent carbohydrate, and 16 percent protein. Results
   showed mean serum cholesterol decreased significantly (-10.8
   percent, p = 0.001) in the oat group compared to the wheat
   group (-4.7 percent). However, the baseline cholesterol was
   higher among the oat bran group. The wheat bran group also
   experienced significantly decreased serum cholesterol compared
   to their baseline (p <0.001). LDL-cholesterol decreased
   significantly compared to baseline for both groups (p = 0.03
   for the wheat bran group and p <0.001 for the oat group). The
   short test period of 2 weeks in this study makes
   interpretation of the results difficult.
       Ranhotra and coworkers (Ref. 122) studied lipidemic
   responses in 17 hypercholesterolemic men consuming foods high
   in soluble fiber. This was a 6-week intervention study with a
   6-week control period prior to the test period. Subjects
   consumed their usual diet during the control period and kept
   daily records of intake for 4 weeks. Subjects were then given
   a list of foods that were identified as good sources of
   soluble fiber and a diet supplement containing 30 g each of
   rice bran and oat bran, and were instructed to incorporate
   foods on the list into their usual diet. Each subject served
   as his own control. Results showed that only 6 (34 percent) of
   the 17 subjects were responders to the soluble fiber
   intervention. The authors reported that not all subjects
   consumed the supplement daily and that intakes of soluble
   fiber varied greatly among the participants. Serum cholesterol
   values decreased from 1 percent to 17 percent compared to
   individual control levels in those responding to soluble
   fiber. However, the authors did not perform a statistical
   analysis of the results, and the results in this study are too
   inconsistent in direction and magnitude to support an effect
   of soluble fiber on serum cholesterol.
       Zhang et al. (Ref. 123) studied the mechanism of
   cholesterol lowering in nine subjects with ileostomies. This
   was a randomized, controlled, cross-over study design.
   Subjects were instructed to consume their own food, which was
   modified to be low fiber, and were assigned to either a low
   fiber diet (supplemented with wheat-flour bread) or a high
   fiber diet (supplemented with oatbran bread) for 3 weeks
   followed by cross-over to the other fiber bread for another 3
   weeks. Ileostomy effluents were collected on sampling days in
   both dietary periods. Subjects were also divided into two
   subgroups according to the amount of bile acids excreted in
   the ileostomy effluents. Results showed subjects with low bile
   acid excretion had significantly increased daily excretion of
   total bile acids on the high fiber diet as compared to the low
   fiber diet. There were no significant differences in daily
   excretion of total bile acids between the high fiber period
   and the low fiber period in subjects with high daily bile
   acilesterol and LDL-cholesterol decreased significantly (p =
   0.01 and p = 0.05, respectively) in all nine subjects on the
   high fiber diet. Subgroup analysis showed that subjects with
   low daily bile acid excretions had significantly reduced serum
   cholesterol on the high fiber diet than on the low fiber diet.
   Subjects with a high daily excretion of bile acids showed no
   significant difference in serum cholesterol between the test
   periods. Conclusions about fiber mechanisms in lowering serum
   lipids in subjects with ileostomies may not apply to the
   general population.
   2. Hypercholesterolemics: Step 1 or 2 diets
       A study by Israelsson et al. (Ref. 86) was submitted with
   a comment. This placebo-controlled, double-blind, cross-over
   intervention study used 30 g/day beet fiber or bread.
   Twenty-seven hypercholesterolemic (serum cholesterol 263 to
   297 mg/dL) women, 55 to 56 years old, were chosen from a CVD
   screening program. Subjects consumed a moderate cholesterol,
   low fat diet with increased polyunsaturated fatty
   acids/saturated fatty acids (PUFA/SFA) for a 1-month run-in
   period and were then randomized to thefiber group or placebo
   for 1 month followed by cross-over to the other diet. The beet
   fiber provided 6 g soluble fiber, 16.5 g insoluble fiber, and
   22.5 g total dietary fiber. Subjects decreased alcohol intake
   after the run-in period. No data on saturated fat or soluble
   fiber intakes were provided. Results showed a significant
   reduction in serum cholesterol in the fiber group compared to
   the placebo after 2 weeks, but not after 4 weeks.
   LDL-cholesterol showed a modest but significant reduction
   after 4 weeks of fiber intervention. HDL remained constant or
   increased significantly after 1-month intervention. The ratio
   of LDL:HDL was significantly reduced at the end of the test
   period. The short test period of this study does not address
   long-term effectiveness of beet fiber.
       Bremer et al. (Ref. 91) evaluated the cholesterol lowering
   effects of oat bran and wheat bran in a randomized,
   single-blind, cross-over, placebo-controlled intervention
   trial. The fibers were incorporated into breads. Twelve
   hyperlipidemic men and women (total serum cholesterol 220 to
   348 mg/dL) were stabilized on a American Heart Association
   (AHA) phase II diet (total fat 25 to 30 percent of energy,
   saturated fat <8 percent of energy, polyunsaturated fat 5 to
   10 percent of energy, cholesterol <250mg/day) for 3 months
   prior to intervention. There was a 2-week run-in prior to test
   during which subjects added additional bread to their diets.
   Subjects were randomized to one of the two fiber groups for 4
   weeks, followed by 2-week wash-out, then cross-over to other
   fiber group. The bread (10 to 12 slices/day) was added to the
   diet in place of other carbohydrate foods. Subjects had a mean
   intake of 44.6 g/day of oat bran (range of 34.2 to 68.4
   g/day). Total dietary fiber intake during the oat period was
   32.2 g, and 34.1 g during the wheat period. Results showed no
   significant differences in total serum cholesterol or
   LDL-cholesterol between the oat and wheat periods. Authors
   account for the lack of observed response on serum cholesterol
   from oat branas due to the lower soluble fiber content of New
   Zealand oat bran compared to oat bran used in other studies.
       Anderson et al. (Ref. 92) evaluated the
   hypocholesterolemic effects of two bulk laxatives, relative to
   psyllium and a placebo (cellulose), in mild to moderately high
   hypercholesterolemic (total serum cholesterol 200 to 300
   mg/dL) men and women. The laxatives were evaluated at the
   manufacturer's recommended dosages. Of the 163 subjects
   screened, 105 completed the 16-week study. Subjects were
   instructed in and consumed the AHA Step 1 diet (total fat 30
   to 33 percent of energy, saturated fat 10 percent of energy,
   carbohydrate 46 percent of energy, cholesterol <300 mg/day)
   for 8 weeks followed by an 8-week parallel treatment with
   diets supplemented with one of three fiber supplements or
   placebo. Fiber sources used were the following bulk laxatives:
   psyl0.2 g/day), calcium polycarbophil (4 g/day), and cellulose
   placebo (4 g/day). The authors note that psyllium and
   methylcellulose were most effective in lowering serum
   cholesterol. There was no significant difference between
   psyllium and methylcellulose in lowering serum cholesterol.
   Soluble fiber was not controlled in this study. Subjects on
   psyllium had the highest soluble fiber intake. Side effects
   were reported for each laxative.
       An unpublished manuscript entitled ``High soluble fiber
   foods reduce serum lipids even when diets are already low in
   saturated fat and cholesterol'' (Ref. 93) was received as part
   of a comment. This 4-week, cross-over, metabolic study in 12
   hyperlipidemic (mean total cholesterol of 272 mg/dL) subjects
   compared a psyllium cereal/low fat diet (9.35 g psyllium/day =
   >3 oz of psyllium cereal) with a similar diet substituting
   wheat bran for the psyllium cereal. The low fat diets were the
   same in both phases of the study, low in saturated fat (<4
   percent of energy) and cholesterol (<50 mg/day) and high in
   carbohydrate ( 60 percent of energy). The psyllium cereal was
   significantly more effective in lowering total and
   LDL-cholesterol than the wheat bran cereal. Mean total
   cholesterol reduction was from 272 mg/dL to 249 mg/dL, and
   from 192 mg/dL 172 mg/dL for LDL cholesterol. Preliminary
   subgroup analysis by the authors of the study suggested that
   patients with both elevated cholesterol and triglycerides
   (Type IIb) showed no reduction in LDL cholesterol, while
   patients with only Type IIa (isolated elevation of
   cholesterol) benefited. Additional concerns are raised by this
   study regarding the usefulness of psyllium in Type IIb
   patients.
       An unpublished manuscript entitled ``High fiber foods
   reduce serum lipids even on diets low in saturated fat and
   cholesterol'' (Ref. 94) was received as a comment. This was a
   cross-over study with 11 hyperlipidemic volunteers. Each
   16-week test period was separated by a 2-month wash-out period
   during which subjects consumed only the Step 2 diet (total fat
   <20 of energy, saturated fat <7 percent of energy, cholesterol
   <100 mg/day, and carbohydrate  60 percent of energy). During
   one metabolic phase, subjects were fed foods considered good
   sources of soluble fiber (e.g., legumes and
   psyllium-containing cereals) as part of the Step 2 diet.
   During the second phase, wheat bran-containing foods were fed.
   Results showed that both the soluble- and insoluble-fiber
   groups lost weight during the 4-month test period. The
   insoluble-fiber group lost significantly more than the
   soluble-fiber group. Although, blood lipids fell on both
   diets, total cholesterol, LDL- and HDL-cholesterol values were
   significantly lower (6.3 percent, 8.6 percent, and 5.7
   percent, respectively) in the soluble-fiber group than in the
   insoluble-fiber group. The soluble fiber diet emphasized foods
   shown in other studies to reduce serum cholesterol, i.e.,
   dried beans, peas, other legumes, oat bran, and a
   psyllium-containing cereal. The actual difference in soluble
   fiber content between the soluble and insoluble fiber diets
   was only about 3.2 g/day, on average. The authors felt that
   the specific foods they fed contribute to lowering of serum
   cholesterol, but expressed concern that not all soluble fibers
   show this effect, and that no mechanism of action is apparent.
   They, therefore, expressed ``concern over lipid lowering
   claims for direct dietary fibers.''
       An intervention study by Anderson et al. (Ref. 95) with 44
   hyperlipidemic (serum cholesterol 200 to 300 mg/dL) men and
   women was conducted using a randomized, double-blind, and
   parallel design. Subjects consumed a Step 1 diet and 3.7 ounce
   (oz)/day (on average) of either a wheat bran cereal or a
   psyllium cereal for 6 weeks. The psyllium cereal provided 10.7
   g/day of psyllium. The psyllium group had a lower total
   cholesterol by about 8 percent and LDL cholesterol by nearly
   13 percent by the end of thebout 1 (lb) pound. Because this
   study had a short test period, it did not address the
   long-term usefulness of psyllium in reducing serum
   cholesterol. In a parallel design, nonblinded clinical trial
   with 59 men and women with total cholesterol between 215 and
   396 mg/dL, Neal and Balm (Ref. 98) placed all subjects on Step
   1 diets for 7 weeks. The control group continued on the Step 1
   diet, and the test group received 20.4 g of psyllium per day
   in the form of Metamucil immediately after breakfast and the
   evening meal for 13 weeks. After the treatment period, the
   psyllium group had a decrease in total cholesterol of 7.1
   percent, while the control group had a 1.6 percent decrease.
   The difference between the test and control groups was 5.5
   percent, a significant decrease. Although there was a 5.1
   percent decrease in LDL in the psyllium group compared to the
   control, this decrease was not significant. The authors failed
   to report the amounts of total fat, saturated fat, and total
   soluble dietary fiber consumed during each period.
       Two short-term studies by Wolever et al. (Refs. 101 and
   102) evaluated the effectiveness of psyllium in lowering serum
   cholesterol (Ref. 101) and its effectiveness when psyllium was
   taken with meals or between meals (Ref. 102). These studies
   were done with men and women, some of whom were on
   lipid-lowering drugs, who were instructed on a Step 2 diet.
   The test periods were for 2 weeks. In both studies, serum
   cholesterol was lowered significantly at the end of the 2
   weeks.
       O'Connor et al. (Ref. 103) conducted a well-controlled
   multicenter, double-blind randomized, parallel group,
   placebo-controlled trial with men and women between the ages
   of 18 and 70 years with a diagnosis of mild to moderate
   primary hypercholesterolemia (see Table 1). A five fiber
   supplement, containing guar and pectin and providing 7.5 g of
   soluble fiber and 2.5 g of insoluble fiber, was administered
   either once or twice a day for 15 weeks with a Step 1 diet.
   The placebo group received 5.2 g of insoluble fiber with no
   soluble fiber before breakfast and dinner. All nutrients were
   kept constant except for fiber. Serum cholesterol and
   LDL-cholesterol were significantly reduced compared to placebo
   in all studies (see Table 1, studies B301 and B302 in). An
   extension of this study evaluated the long-term usefulness of
   the five fiber supplement for an additional 36 weeks (total of
   51 weeks). Significantly reduced levels of total and
   LDL-cholesterol were maintained (5.3 percent and 8.4 percent,
   respectively) compared to baseline. This study shows both the
   ability of a particular soluble fiber product for reducing
   blood lipids and the long-term benefits of soluble fiber
   supplementation with a low fat diet.
       An unpublished study (Ref. 108) evaluated the
   hypocholesterolemic effect of psyllium in 23
   hypercholesterolemic men (mean total cholesterol greater than
   240 mg/dL). Using a double-blind, double cross-over design,
   the subjects were randomly assigned to either the
   psyllium-wheat bran-psyllium group or to the wheat
   bran-psyllium-wheat bran group for 8, 5, and 5 weeks,
   respectively. Subjects consumed a total of 10 g of soluble
   fiber/day from psyllium and 2 g of soluble fiber per day from
   wheat bran. All subjects consumed a Step 1 low fat diet as the
   base diet. Results showed significant cholesterol and
   LDL-cholesterol lowering (4.3 percent) with psyllium compared
   to the wheat phase of the test. Initial cholesterol values for
   each group were not given.
       A 1987 study by Turnbull and Leeds (Ref. 111) was
   submitted with comments from the food industry as further
   evidence for a hypocholesterolemic effect of soluble fiber
   from oat bran. In this study, 17 free living men and women
   aged 23 to 73 years (serum cholesterol levels above 232 mg/dL)
   were given varying degrees of instruction on a low fat diet
   (<35 percent calories from fat). All subjects followed this
   diet for a 1-month run-in period. Eight subjects were followed
   intensively during the run-in pere then randomly assigned to
   receive either 150 g/day oats (from cereal and muffins) or 100
   g/day wheat flour biscuits for 1 month followed by cross-over
   to the other fiber diet for an additional month. The oat
   products provided 5.4 g of soluble fiber per day. The wheat
   products provided 3.1 g/day. Results during the run-in period
   showed a significant loss of weight and reduction in serum
   cholesterol in the group of eight subjects studied intensively
   during the first month. At the start of the first test period
   (whether oat or wheat) all subjects had a mean fat intake of
   34 percent of calories. During the oat period, the subjects'
   energy intakes increased and fat intake increased to 35
   percent. Mean body weight, however, remained constant.
   Combining all results from the oat-wheat and wheat-oat
   periods, serum- and LDL-cholesterol fell significantly (p =
   0.02 and p = 0.003, respectively). During the wheat period,
   the subjects' caloric intake increased with a mean body weight
   increase of 0.3 kilogram (kg). Total fat consumption during
   the wheat period was 36 percent of calories. Mean serum
   cholesterol of the combined period showed a nonsignificant
   increase (1.6 percent) in serum cholesterol during the wheat
   period. LDL-cholesterol showed a nonsignificant increase. In
   the eight subjects followed intensively throughout the study,
   the results showed that most of them had further reductions in
   total and LDL-cholesterol on the oat diet beyond the low fat
   diet alone. In a later study (Ref. 126) of the samedesign and
   using the same levels of oats and wheat flour, the authors
   reported favorable changes in apolipoprotein A1 and no change
   in apolipoprotein B in subjects on the oat diet.
       The purpose of the study by Fukagawa et al. (Ref. 114) was
   to evaluate the effects of a very high carbohydrate high fiber
   (HCF) diet on peripheral-tissue insulin responsiveness in a
   group of healthy young men (ages 18 to 24 years, Group A) and
   in a group of older men and women (ages 67 to 86 years, Group
   B). The young men had normal serum cholesterol (199 mg/dL) and
   the older adults were hypercholesterolemic (237 mg/dL). The
   subjects were studied while consuming their usual ad libitum
   diet and after consuming a HCF diet for 21 to 28 days. The
   older group was admitted to the metabolic ward for the
   duration of the study. The younger subjects only ate their
   meals on the metabolic ward. The study was not blinded or
   placebo controlled. Test diets provided the following: Group
   A-- 23.6 g/day soluble fiber, 88.2 g of plant fiber/day, 134
   mg dietary cholesterol, 14 percent energy as fat, 3 percent
   energy as saturated fat, 69 percent of energy as
   carbohydrates; Group B-- 17 g of soluble fiber, 67.7 g of
   plant fiber/day, 90 g of dietary cholesterol, 15 percent
   energy as fat, 3 percent energy as saturated fat, and 70
   percent energy as carbohydrates. Results showed a significant
   reduction in serum cholesterol in both groups after 4 weeks
   (Group A:  26 percent; Group B: 45 percent). The results of
   this study are inconclusive for an effect of fiber on serum
   cholesterol (which was not the objective of the study) because
   the subjects' ad libitum diets were significantly higher than
   the test diet in fat (37 to 42percent of calories compared to
   approximately 14 percent on the test diet), saturated fat (15
   to 17 percent of calories versus 3 percent of calories), and
   lower in carbohydrates (40 to 45 percent of calories versus 68
   to 70 percent of calories). In addition, there was no control
   group.
       Anderson et al. (Ref. 118) evaluated the
   cholesterol-lowering benefits of psyllium-enriched cereal in
   subjects with mild to moderate hypercholesterolemia (serum
   cholesterol range of 200 to 300 mg/dL). Subjects consumed
   their usual diets for 1 week before being randomly assigned to
   receive psyllium-flake or wheat bran flake cereal for 6 weeks.
   Subjects were also instructed on a Step 1 diet and asked to
   adhere to it for 6 weeks. Soluble fiber intake during the
   treatment period was 5.9 g per day for theowed significantly
   reduced serum- and LDL-cholesterol in the psyllium group
   compared to the wheat group. Serum cholesterol was reduced
   8.36 percent (p = 0.01) and LDL-cholesterol 12.9 percent (p =
   0.01) in the psyllium group. There was no significant change
   in serum- or LDL-cholesterol in the wheat group.
   3. Normocholesterolemics: ``typical'' or ``usual'' diets
       Resnicow et al., 1991 (Ref. 96), measured total serum
   lipids in a population of 31 Seventh Day Adventists, ages 5 to
   46 years, who had consumed a pure vegetarian diet for at least
   6 months prior to taking of blood samples. Diets of vegans
   were compared to those of omnivore controls. Blood samples
   were not taken from the controls and blood values for these
   subjects were derived from the Lipid Research Clinics
   Population Studies Data Book. Results showed that the adult
   vegans consumed significantly less energy and energy from fat
   (31 percent versus 38 percent of calories), total fat,
   saturated fat, monounsaturated fat, cholesterol, and protein.
   They also consumed significantly more fiber (45 g/day versus
   20 g/day) than the omnivores. Total dietary soluble and
   insoluble fibers were not assessed. Foods consumed in greater
   frequency by vegans included almonds, cashews, and their nut
   butters, dried fruits, citrus fruits, soy milk, and greens.
   Total serum cholesterol for vegans was approximately 23
   percent lower (139 mg/dL versus 182 mg/dL) than expected
   values for omnivores.
       One study submitted with comments evaluated the effect of
   glucomannan, a pectin-like gel fiber derived from purified
   tubers of Amorphophallus koniac K. Koch, on serum cholesterol
   and weight reduction in obese patients consuming their normal
   diets (Ref. 117). Weight loss and serum cholesterol decreased
   significantly in the test group compared with the placebo
   group at the end of the 8-week trial. After 4 weeks on the
   test product, subjects had a mean weight loss of 4.9 lb and
   mean serum cholesterol reduction of 20.9 mg/dL. After another
   4 weeks, weight loss was only 0.6 lb and serum cholesterol was
   reduced only 0.8 mg/dL. Because weight loss and serum
   cholesterol are closely correlated, the effect of glucomannan
   on serum cholesterol cannot bedetermined from this study.
       An unpublished study studied the mechanism of serum
   cholesterol reduction by oat bran (Ref. 124). This was a
   2-month metabolically-controlled intervention trial with nine
   normocholesterolemic men. A single isotope was used to
   determine bile acid kinetics during the oat bran period.
   During the first month, subjects consumed a constant diet
   provided in a metabolic unit. The fat content of the diet was
   35 percent of the energy. The total soluble fiber content of
   the low fiber diet ranged from 3.0 to 4.9 g/day and for the
   high fiber period 9 to 12 g/day. During the second month, this
   same diet was supplemented with 100 g of oat bran per day.
   Results showed significantly lowered serum cholesterol during
   both periods. Serum cholesterol was 14 percent lower compared
   to the prestudy period during the low fiber period and 22
   percent lower during the high fiber period. Serum cholesterol
   during the high fiber period was also significantly lower than
   that of the low fiber period (an additional decrease of 9
   percent). Bile acid excretion approximately doubled during the
   high fiber period.
   4. Normocholesterolemics: low fat diets
       Nervi et al. (Ref. 107) reported than an intake of 120
   g/day of legumes for 30 to 35 days significantly lowered serum
   cholesterol (from 162 to 143 mg/dL) in 20 Chilean young men
   compared to responses of men on a control diet. The men
   consumed beans,peas, or lentils each day as part of a diet
   that provided 33 percent of calories from fat and 12.5 g of
   total fiber. The study was designed to evaluate the hypothesis
   that legumes may be a risk factor for cholesterol gallstones
   in certain subpopulations.
   5. Other studies
       Evidence for the cholesterol-lowering effect of soluble
   fiber from oats was evaluated using meta-analysis (Ref. 125).
   In this study, after pooling the raw data from 5 investigators
   who had looked at the effect of consumption of oat products on
   bloodtotal cholesterol, a modest reduction (average decrease
   of 5 to 6 mg/dL) on blood total cholesterol levels was found.
   The decrease in blood total cholesterol was largest in those
   trials with initially higher blood total cholesterol levels,
   particularly where an intervention dose of 3 g or more of
   soluble fiber from oats was used. To assess whether other
   dietary factors, i.e., dilution of saturated fat and calorie
   intakes by the oatmeal or oat bran addition to diets, might
   have been responsible for the drop in blood total cholesterol
   levels, the authors used the experimentally derived,
   predictive equation of Keys to see if dietary changes in fat
   components of the test diets could account for the observed
   decreases in serum cholesterol. The fat and saturated fat
   changes did not appear to be responsible for the drop in serum
   cholesterol levels, thus suggesting that some other factor in
   the test diets (e.g., the soluble fiber fraction) was
   responsible for the observed effects. The authors concluded,
   therefore, that incorporation of oats (a rich source of
   soluble fiber, primarily as [-glucan) into diets causes a
   modest decrease in average blood cholesterol. The authors also
   suggested that there was a dose-response relationship between
   the amount of soluble fiber from oats and the reduction in
   blood cholesterol levels, with intakes of soluble fiber from
   oats above 3 g/day showing more effect than lower intakes.
   Additionally, the authors noted that other components in oats
   may play a role in the observed cholesterol reduction and
   suggested the need for long-term clinical trials (6 months or
   more) with multiple doses to verify their conclusions from the
   meta-analysis.
   6. Summary of human studies
       The human studies reviewed above suffered from many of the
   same design flaws noted in the proposed rule on health claims
   for dietary fiber and cardiovascular disease (56 FR 60582 at
   60591). Some studies were conducted with very small sample
   sizes (Refs. 84, 85, 87, 90, 91, 93, 94, 97, 110, 111, 114,
   122, 123, 124, and 126). Another limitation was short study
   times (Refs. 87, 101, 102, 110, 114, 120, 121, and 123).
   Inadequate control of confounding factors, such as concomitant
   weight losses and changes in other dietary components which
   may have affected results, plagued some studies (Refs. 86, 96,
   107, 108, 110, 117, 119, and 120). The absence of adequate
   dietary intake data to assure that dietary changes other than
   differences in soluble fiber intakes had not occurred was a
   problem for a number of studies (Refs. 84, 86, 87, 88, 90, 91,
   95, 97, 99, 105, 106, 119, and 123).
       Several studies were suggestive of positive effects of
   soluble fiber intakes on blood cholesterol levels. One study
   provided evidence of a relationship between consumption of
   foods high in soluble fiber and reduced levels of blood total-
   and LDL-cholesterol levels (Ref. 118). In a comparison of a
   breakfast cereal fortified with psyllium, a rich source of
   soluble fiber (12 g of psyllium/day from 114 g/day of a
   psyllium-containing cereal) to a wheat bran cereal which
   contributed negligible amounts ofsoluble fiber, the
   psyllium-containing cereal was associated with lower blood
   total- and LDL-cholesterol levels after 6 weeks than were
   observed in subjects following a diet containing wheat bran
   cereal. In another study (Ref. 104), consumption of 123 gof
   oat bran cereal (contributing 10.3 g soluble fiber daily)
   versus consumption of 54 g of wheat bran cereal per day
   (contributing 3.4 g of soluble fiber daily) was associated
   with lower blood total- and LDL-cholesterol levels after 4
   weeks. Dietary intakRefs. 83 and 91). Several explanations for
   the lack of a relationship in these studies were offered by
   the authors, including the possibility that the oat bran used
   was low in soluble fiber content (Refs. 83 and 91). The study
   by Neal and Balm (Ref. 98)showed significantly lower blood
   total-cholesterol levels with consumption of a
   psyllium-fortified cereal, but the decline in LDL-cholesterol
   levels compared to the control was not statistically
   significant. A small body weight loss in both groups may have
   confounded the relationships. However, since it is the
   LDL-cholesterol, rather than the total-cholesterol, that is
   the desired endpoint for evaluating beneficial changes, the
   lack of statistical significance for serum LDL-cholesterol
   levels limits this study's usefulness.
       Finally, the meta-analysis on the cholesterol-lowering
   effect of oat products was useful and suggested a benefit from
   oat consumption. However, the authors noted that, while grams
   of soluble fiber were chosen to represent the dose of oat
   product, itis entirely possible that other components of oats,
   as well as the way in which the oat product is prepared, may
   also play a role in reduction of blood LDL-cholesterol levels.
   The authors recommended that, because there are several
   components of oats which could provide beneficial effects on
   blood cholesterol levels, future clinical studies should test
   multiple doses of oat products with the simultaneous
   measurement of other possible active components, including
   soluble fiber, [-glucan, and tocotrienols.
       The most definitive results linking soluble fiber intakes
   to beneficial changes in blood cholesterol levels were for
   studies in which dietary supplements of guar (Refs. 99 and
   100), gum arabic (Ref. 109), psyllium (Ref. 92), or a
   combination of soluble fiber sources (Ref. 103) were given.
   Some of these studies (Refs. 99, 100, and 109) also, however,
   failed to provide adequate information on dietary intakes,
   thus limiting the ability to rule out possible confounding
   effects from other dietary changes that may have occurred
   concomitantly with addition of these supplements. A series of
   well-conducted clinical trials were done to design and test
   the effectiveness of fiber mixtures (guar, pectin, psyllium,
   and locust bean) on blood cholesterol levels (Refs. 103 and
   106). Early studies in one series (Ref. 106) showed no benefit
   from acacia gum alone or a mixture of acacia gum, psyllium and
   guar (Studies 1 and 2). Only when a mixture of pectin,
   psyllium, guar and locust bean was given were beneficial
   effects seen (Study 4). These results strongly suggest that
   benefits of fiber supplements are not readily predicted by an
   analytical definition of soluble fiber, but rather vary, in
   some unknown way, among different sources or combinations of
   sources, of soluble fiber. Thus, generalizing results from one
   fiber source to another must be done cautiously.
   
   B. Animal studies submitted with comments
   
       FDA received a number of animal studies submitted as
   comments. FDA has reviewed these studies as described below.
       1. Relationship between specific soluble fibers and plasma
   cholesterol
       Ney et al. (Ref. 127) evaluated the effect of soluble oat
   fiber on blood very low density lipoprotein (VLDL), low
   density lipoprotein (LDL), and high density lipoprotein (HDL)
   levels by feeding male rats cholesterol-raising diets (diets
   which contained 1 percent cholesterol and 0.2 percent cholic
   acid as the stimulus for increasing blood cholesterol levels)
   and 6 percent dietary fiber from cellulose (control) or from
   three oat products with increasing levels of soluble fiber:
   Oat bran, high fiber oat flour or a processed oat product.
   Compared to the cholesterol-fed cellulose control, all oat
   fibers lowered plasma total cholesterol by 25 to 45 percent,
   lowered VLDL + LDL cholesterol levels by 40 to 60 percent, and
   raised HDL-cholesterol by 25 to 40 pt, which contained 40
   percent more soluble fiber than oat bran or oat flour,
   resulted in a lipoprotein profile similar to that obtained
   without the cholesterol-raising stimulus of dietary
   cholesterol and cholic acid. The oat product with the highest
   soluble fiber content was, therefore, more effective with
   ingestion of the cholesterol-raising diet than was the oat
   bran or oat flour with lower amounts of soluble fiber. The
   authors suggested that these data, which show greater benefits
   as the soluble fiber content increases, support the suggestion
   that soluble fiber is the component of oat fiber responsible
   for the cholesterol-lowering effect of oat-containing diets.
       Nishini et al. (Ref. 128) evaluated the effect of dietary
   fibers from oat bran, wheat bran, cellulose, and pectin on
   plasma lipoproteins, apolipoproteins and enzymes involved on
   cholesterol metabolism in non-fasted rats. The animals were
   fed experimental diets estimated to contain 8 percent dietary
   fiber by weight. Results showed that pectin-fed animals (i.e.,
   animals receiving the highest soluble fiber) had significantly
   lower serum cholesterol, HDL-cholesterol, and apolipoprotein
   A-1 levels compared to the fiber-free control. Total
   cholesterol levels in the wheat-bran-fed (primarily insoluble
   fiber) or oat bran-fed (mixed soluble/insoluble fiber) animals
   were not significantly different from the fiber-free control.
   No data were given on the soluble fiber content of the diets,
   although the pectin diet would be expected to have the highest
   level of soluble fiber. Results showed that blood lipid
   distributions are affected differently by dietary fibers, and
   that changes among lipid components frequently occurred
   without a change in overall cholesterol concentrations.
       Prentice et al. (Ref. 129) compared the effects of ground
   and rolled caryopses of barley and rolled oats to ground corn
   on hepatic cholesterol and fatty acid synthesis in chickens.
   Approximately 7 to 8 percent of the barley and oat cereals was
   [-glucan; corn had less than 1 percent [-glucan. Both barley
   and oats decreased plasma total cholesterol by 32 percent and
   25 percent, respectively. The authors attributed the effect on
   serum cholesterol to the higher soluble fiber content of
   barleyand oat diets.
       Summaries of unpublished animal studies (Ref. 130)
   suggested that oats may be effective in lowering plasma
   cholesterol compared to controls fed white flour, wheat flour,
   or corn starch. However, the data were preliminary and, thus,
   had limited usefulness.
       Other summaries of unpublished animal studies (Ref. 136)
   suggested that rolled oats (75 percent by weight in diet)
   significantly lowered serum cholesterol in chicks. Extruded
   oat bran, equivalent to 47 percent of oat bran by weight in
   the diet, significantly depressed (p = 0.05) serum cholesterol
   in chicks. Oat gum at 0.05 percent and 0.10 percent by weight
   in the diets of rats significantly lowered serum cholesterol.
   Data comparing several oat fractions fed to chicks suggested
   that oat gum may be the active cholesterol depressant
   component, and that ground rolled oats and instant oatmeal are
   equal in effect to the defatted, defibered oat flour. The oat
   oil also had a depressive effect. Additionally, pectin (high
   soluble fiber) was found to significantly lower serum
   cholesterol in rats. These studies were done in the mid-1970's
   and did not have analyzed values for soluble fiber content of
   the respective diets. Methods and data were not well
   described, making results difficult to interpret.
       Qureshi et al. (Ref. 131) reported the effects of diets
   supplemented with either corn (61.5 percent by weight of the
   base diet), wheat (75 percent of the diet), barley (73.5
   percent), oats (74.5 percent), or rye (73.5 percent) on serum
   cholesterol in chickens. This was part of a study to
   investigate the effects of dietary cereals on the metabolic
   regulation of lipid metabolism in chicken livers. Compared to
   corn, barley and oats lowered serum cholesterol 45 percent and
   32 percent, respectively. Presuigher in soluble fiber content
   than was corn. However, no fiber content data were presented.
   Additionally, body weights for animals consuming the barley
   and oats diets varied, making it difficult to ascribe the
   effects to fiber per se.
       Ranhotra and co-workers (Ref. 132) evaluated the effect of
   oat bran and oat bran concentrate on serum lipids in rats.
   Animals were fed experimental diets containing oat bran (5.57
   percent soluble fiber) or oat bran concentrate (13.75 percent
   solublefiber). Results showed the high soluble fiber content
   of the oat bran concentrate was associated with a
   significantly lower serum cholesterol level, but that the oat
   bran diet (which contained lower amounts of soluble fiber) and
   the diet free of solublefiber were not associated with lower
   serum cholesterol levels.
       Shinnick et al. (Ref. 133) evaluated the ability of
   various sources and forms of oat fiber to lower plasma and
   liver cholesterol in male rats fed a diet with 6 percent
   dietary fiber as cellulose (0 percent soluble fiber), oat bran
   (7 percent solublefiber), high fiber oat flour (8 to 10
   percent soluble fiber), or one of four processed high fiber
   oat flours (8 to 12 percent soluble fiber). All diets were
   supplemented with 1 percent cholesterol and 0.2 percent cholic
   acid to stimulate increased levels of blood total cholesterol.
   Results showed that all oat products significantly lowered
   serum cholesterol compared to the control. In a second
   experiment, diets containing 4 percent dietary soluble fiber
   in a processed oat flour significantly lowered serum
   cholesterol. The processed oat flours had higher soluble fiber
   fractions than the less processed oat products.
       In a study published in 1983, Rogel and Vohra (Ref. 135)
   reported no effect from five varieties of oats (oat bran or
   oat hulls) fed for 4 weeks on the serum cholesterol levels of
   quail. No data on soluble fiber were given.
       Kritchevsky et al. (Ref. 138) evaluated the
   cholesterol-raising effects of oat and wheat bran on blood
   cholesterol levels in rats in a three-week feeding study in
   which semipurified diets containing 0.5 percent cholesterol
   and 10 percent oat bran, wheat bran or cellulose were fed to
   male rats. Weight gains varied among the diets. Results showed
   that, under these study conditions, there were no effects of
   diet on any of the serum lipids. No data on soluble fiber
   content of the diets were provided.
       Kahlon et al. (Ref 140) evaluated the effects of rice
   brans (full-fat or defatted), oat bran, or rice-wheat bran
   combinations on cholesterol in hamsters. The control diet
   contained 10 percent cellulose and 0.5 percent cholesterol.
   Test diets were composed of the control diet plus one of the
   brans. The oat bran contained 8 percent dietary soluble fiber
   versus 2 to 3 percent soluble fiber in the other brans.
   Results showed that rice bran (with fat) and oat bran resulted
   in significantly lower plasma cholesterol than the control
   diet. Defatting rice bran resulted in loss of its
   cholesterol-lowering properties, suggesting that it is the
   lipid portion rather than the fiber portion of rice bran which
   is responsible for its cholesterol-lowering effects.
       In another study, rats were fed a diet containing 1
   percent cholesterol and 0.2 percent cholic acid with added
   cellulose, oat gum, chitosan, or cholestyramine (5 percent of
   the diet). Cholestyramine, oat gum, and chitosan all
   significantly lowered blood and liver cholesterol levels, with
   the greatest effect with cholestyramine, a commonly used
   cholesterol-lowering drug. Oat gum, at 5 percent of the diet,
   reduced serum cholesterol by 23 percent (Ref. 141).
       Forsythe et al. (Ref 142), in a study published in 1978,
   evaluated the influence of source and particle size of dietary
   fibers on hypocholesterolemic effects in rats. The sources of
   fiber were cellulose, wheat bran, wheat midlings, oat bran,
   oat flourto the fiber-free group.
       Significantly lower values of plasma total cholesterol and
   liver cholesterol in rats fed fiber sources known to contain
   soluble fibers (oat bran, oat gum, and pectin) compared to
   rats fed cellulose (insoluble fiber) were reported by Chen et
   al. (Ref. 143). Rats were fed a base diet containing 1 percent
   cholesterol and 10 percent by weight of one of the four fiber
   types. The greater effect of oat gum compared to oat bran was
   interpreted by the authors to suggest that the plasma and
   liver cholesterol-lowering effects of oat bran are due to its
   gum fraction.
       Chen and Anderson (Ref. 144) examined the effects of
   fiber-supplemented diets on total and LDL-cholesterol in rats.
   Rats were fed one of five experimental diets containing
   sucrose and 10 percent plant fibers. The diets were as
   follows: sucrose and cellulose, sucrose-cholesterol with
   cellulose, sucrose-cholesterol-pectin,
   sucrose-cholesterol-guar gum, and sucrose-cholesterol-oat
   bran. Results showed that fiber-fed rats had significantly
   lowered plasma cholesterol than rats that received cellulose;
   the lowest concentrations were observed in the pectin-treated
   group. No data were given on the soluble fiber content of the
   test diets, although the pectin would presumably contain the
   highest concentration of soluble fiber.
       Chen and Anderson (Ref. 146), in two experiments, examined
   the effects of guar gum or wheat gum on the plasma and liver
   lipid levels of rats. In the first experiment, animals were
   fed one of three diets for three weeks: diet plus sucrose and
   10 percent cellulose; sucrose, 15 percent wheat bran, and 4
   percent cellulose; and sucrose and 45 percent wheat bran. Each
   diet provided 10 g of plant fiber. In experiment two, animals
   were fed one of four diets: diet plus sucrose, sucrose plus
   cholesterol, sucrose plus cholesterol and wheat bran, or
   sucrose, cholesterol, and guar gum. The wheat and guar diets
   provided 7 g of plant fiber; the other two diets provided 4 g
   of plant fiber. Results from experiment 1 showed that the two
   wheat bran diets (high in insoluble fiber) significantly
   raised plasma cholesterol compared to the control. Results of
   experiment 2 showed the guar diet (high in soluble fiber)
   significantly lowered serum cholesterol compared to those fed
   the sucrose-cholesterol or sucrose-cholesterol-wheat bran
   diets. Plasma cholesterol was similar between the
   sucrose-cholesterol and wheat groups.
       Wilson et al. (Ref. 148) examined the influence of
   different soluble fibers upon the metabolism of lipids in
   genetically hyperlipemic, obese Zucker rats. Four diets were
   tested: a non-fiber diet (no added fiber), a 0 percent soluble
   fiber diet (cellulose fiber), an oat bran diet (33 percent
   soluble fiber), and a pectin-citrus fiber diet (100 percent
   soluble fiber). Except for the control, the fibers provided 10
   percent of the total diet weight. Results showed
   non-significant decreases in total plasma cholesterol in the
   oat bran and pectin groups compared to the no-fiber and
   cellulose groups. However, the pectin group (the highest
   soluble fiber group) had changes in blood lipid components
   associated with reduced risk of heart disease:
   significantlylower LDL-cholesterol and significantly higher
   HDL-cholesterol.
       Welch et al. (Ref. 149) evaluated the hypocholesterolemic
   effects of oat bran fractions. Oat bran was separated into
   five fractions: oil, insoluble fraction (rich in starch and
   insoluble fibre), protein-rich, oat gum, and soluble residue.
   These werefed to chicks. Results showed that, compared to the
   control, only the diets containing oat bran, oat gum, or the
   protein fractions significantly lowered plasma cholesterol.
   Oat gum was the only fraction which had the same effect in
   reducing plasma cholesterol levels as did the native oat bran.
   Thus, oat gum was concluded to be the main
   cholesterol-lowering component of oat bran. Beta-glucan was
   the main component of the gum fraction. However, the protein
   fraction also had a beneficial effect on plasma chportion.
       Five sources of dietary fiber were compared for their
   effect on blood and liver cholesterol in beef-fed C57BL/6 male
   mice (Ref. 150). Mice were fed one of the following fiber
   supplements (7 percent dietary fiber) with the experimental
   diet: soybean fiber, rice bran (full fat), oat bran, barley
   bran, and mixed bran (one-third each of rice, barley, and oat
   brans). Results showed significantly lower plasma total
   cholesterol compared to the control in the rice bran and
   soybean fiber groups.
       Ginter et al. (Ref. 151) reported that addition of 5
   percent citrus pectin (a high soluble fiber source) and 0.5
   percent ascorbic acid to a high-fat diet fed to guinea pigs
   prevented cholesterol accumulation in blood serum and the
   liver.
       Kakis et al. (Ref. 152) reported reduced serum cholesterol
   and HDL levels in all psyllium-fed (a soluble fiber) animals
   as compared to wheat bran-fed (a relatively high insoluble
   fiber source) animals after a three week experimental period.
   Relativeto the appropriate wheat bran control, psyllium had a
   graded serum hypocholesterolemic effect that varied from a
   high of 41 percent at low dietary cholesterol concentrations
   to 26 percent at high dietary cholesterol concentrations.
   However, HDL cholesterol (in mg percent and as a percent of
   total cholesterol), the ``beneficial blood cholesterol,'' was
   lower in the psyllium groups than in the respective wheat bran
   controls. Thus, the overall benefit of psyllium was not clear
   from this study.
       Life Sciences Research Office's (LSRO) 1982 report to FDA
   on the health aspects of psyllium seed and other food
   ingredients (Ref. 153) was submitted with comments. Summaries
   of studies with beagles fed psyllium-supplemented diets
   consistently showedlowered serum cholesterol compared to
   controls.
   2. Animal studies: dose-response relationship between soluble
   fiber and plasma cholesterol
       Shinnick et al. (Ref. 134), following evaluation of
   several approaches to improve the cholesterol-fed rat model
   used to evaluate the hypocholesterolemic potential of foods,
   fed nine levels of a high fiber oat flour (HFOF) derived from
   oat bran to male Sprague-Dawley rats. Ingestion of increasing
   amounts of HFOF, containing 0 to 10 percent dietary fiber, by
   rats in which high blood cholesterol levels had been produced
   with 1 percent cholesterol and 0.1 percent cholic acid,
   resulted in a significant inverse relationship between serum
   cholesterol levels and HFOF intake for serum and liver
   cholesterol levels. Similar results were obtained for liver
   cholesterol levels. The authors suggested that, although this
   study does not distinguish among the components of HFOF that
   may contribute to the cholesterol-lowering effect, the
   observed dose-response relationship in the rat model is
   suggestive that larger intakes of soluble oat fiber sources
   may be accompanied by greater reductions in serum cholesterol
   levels in humans.
       Turley and co-workers (Ref. 147) conducted a dose-response
   study to three levels of psyllium supplementation in the diets
   of hamsters. Results were compared to two other nonabsorbable
   polymers known to have cholesterol-lowering effects (i.e.,
   cholestyramine and surfomer). Animals were fed diets
   containing 0 percent, 1 percent, 4 percent, or 7.5 percent
   psyllium. Results showed significantly lowered plasma
   cholesterol compared to the control group in animals consuming
   4 percent and 7.5 percent psyllium. The group consuming 7.5
   percent psyllium had the lowest plasma cholesterol, although
   the authors did not report any significant difference between
   the 4 percent and 7.5 percent psyllium groups. While all three
   polymers were effective in reducing plasma total and
   LDL-cholesterol levels, each exerted different quantitative
   and qualitative effects on bile acid and cholesterol
   metabolism, suggesting that mechanisms of action may vary by
   fiber type.
   3. Animal studies: relationship between [-glucan and plasma
   cholesterol
       Three cultivars of hull-less barley containing [-glucans
   were evaluated for their hypocholesterolemic responses in
   chickens (Ref. 137). The authors identified the Arizona
   cultivar of barley as a waxy-starch genotype of high molecular
   weight andwith a high [-glucan content. The Washonupana
   cultivar was second highest in molecular weight and is also
   described as a waxy-starch type genotype. The third cultivar,
   Franubet, has the lowest molecular weight and is not waxy in
   texture. These latter two genotypes have normal [-glucan
   contents. Both the Arizona and Washonupana varieties produce
   highly viscous slurries in water, and this viscosity was
   greatly reduced by addition of the enzyme, endo-[-glucanase,
   which destroys the [-glucan. Results of feeding studies in
   rats showed that both the Arizona and Washonupana cultivars
   reduce serum cholesterol in chickens. The Franubet variety had
   no effect. The authors interpreted these results to mean that
   the cholesterol-lowering properties of the Arizona and
   Washonupana cultivars were probably a function of their
   viscous [-glucan content.
       Klopfenstein and Hoseney (Ref. 139) evaluated the
   cholesterol-lowering effect of [-glucan-enriched bread. Oat,
   barley, wheat, and sorghum breads were made with and without
   [-glucan from each type of grain (e.g., oat [-glucan was
   processed into oat bread) and fed to rats for 35 days. Breads
   containing [-glucans from oats and barley adversely affected
   weight gains and feed efficiencies. Results showed lowered
   serum and liver cholesterol in rats fed the [-glucan-enriched
   breadthan those fed the control breads. Given differences in
   weight gains, results are difficult to interpret.
       Fadel et al. (Ref. 145) evaluated the hypocholesterolemic
   effects of [-glucans in different barley diets fed to chickens
   and the influence of [-glucanase on the hypocholesterolemic
   effects. The animals were divided into five groups and fedone
   of five diets: a corn-diet, Washonupana (WSNP) barley, WSNP
   with [-glucanase, Fraubet (FNBT) barley, FNBT with
   [-glucanase. Results showed significantly lowered serum
   cholesterol only in chicks fed the WSNP diet. LDL-cholesterol
   levels forall barley fed chicks were significantly lower than
   that of the corn diet-fed chicks. Only the WSNP fed chicks had
   serum cholesterol significantly lower than WSNP fed chicks
   with supplemental [-glucanase. There was no significant
   difference in total cholesterol in FNBT and FNBT plus
   supplemental [-glucanase fed chicks. [-glucan in the
   Washonupana barley has hypocholesterolemic effects and
   addition of the [-glucanase to this diet reverses the
   hypocholesterolemic effects. However, thelack of similar
   finding with added [-glucanase to the Franubet variety
   suggests that [-glucans may differ among different varieties.
   4. Summary of results from animal studies
       The animal studies received as comments, and reviewed
   above, provide evidence to support the likely effectiveness of
   soluble fibers relative to the cholesterol-lowering
   characteristics of diets high in some cereals. However, the
   animal studies, likethe human studies, fail to provide
   adequate specifications characterizing the test fiber sources.
   Indeed, similarly to many human studies, many animals studies
   did not analyze experimental diets for soluble fiber content,
   nor provide descriptions of thephysical characteristics or
   commercial sources of the soluble fibers used as test
   substances.
       Results from the animal studies showing effectiveness of
   increasing levels of oat products are suggestive, but not
   conclusive, evidence of a dose response for soluble fiber
   (Refs. 127, 132, 134). A specificity for the gum portion of
   oats, the major source of soluble fiber in oats--and
   specifically, of beta-glucan as the predominant soluble fiber
   source--is suggested by several studies (Refs. 127, 129, 132,
   134, 137, 141, 143, 145, and 149). However, the solubility
   (and possibly, the effectiveness) of bef. 154). These results,
   therefore, suggest that analysis of the beta-glucan content
   per se may not be sufficient to characterize the active
   soluble fiber content of foods.
       Other soluble fibers were also shown to have serum
   cholesterol-lowering effects in animals, including pectins
   (Ref. 128, 143, 144, 148, and 151), psyllium (Refs. 147 and
   152), and guar gum (Ref. 146). However, in the psyllium study
   (Ref. 152), adverse effects on HDL-cholesterol levels were
   observed, raising the issue that individual fibers need to be
   evaluated as to their overall effect on all relevant blood
   lipid components, not simply limited to blood total or
   LDL-cholesterol levels. The various fiber sources also appear
   to have different mechanisms of action and different relative
   magnitudes of effect (Refs. 128 and 147), thus suggesting that
   caution is necessary before generalizing from one type of
   dietary fiber to another.
   
   C. Conclusions from new studies
   
       FDA reviewed over 40 human studies that became available
   since publication of its proposed rule, and a number of animal
   studies submitted as comments. The most commonly studied
   soluble fiber sources were oat bran, pectin, guar gum, and
   psyllium. Other fibers studied were wheat germ, beet fiber,
   and gum arabic. A few studies evaluated the effects of
   mixtures of soluble and insoluble fibers or food sources of
   soluble fibers.
       FDA evaluated results from these studies in light of
   studies reviewed in the proposal and conclusions from Federal
   government and other authoritative reviews. In the proposal,
   FDA noted that, although most reviews by authoritative
   scientific bodies had concluded that diets rich in
   water-soluble fiber fractions were associated with
   cholesterol-lowering effects in humans, it was not possible to
   conclude that the observed effects were due to the soluble
   fiber or to other components associated with consumption of
   foods rich in soluble fiber. FDA also noted that there was
   some evidence that different types of soluble fiber have
   different effects, and that the analytical measure of soluble
   fiber may not be adequately predictive of its physiological
   effects. Thus, FDA concluded that:
       Overall, the available data are not sufficient to
   demonstrate that it is the total soluble dietary fiber, or a
   specific measurable and quantifiable subcomponent, that is
   related to lower blood cholesterol levels.
   (56 FR 60582 at 60592).
       The newer evidence available since the proposed rule and
   reviewed above do not change the conclusions reached by the
   earlier review. If, however, additional information becomes
   available to demonstrate that a specific soluble
   fiber-containing product, a soluble fiber-containing
   ingredient, or a more highly specified form of foods soluble
   fiber is effective in lowering bloood LDL-cholesterol and does
   not adversely affect other risk factors for cardiovascular
   disease (e.g., blood HDL-cholesterol levels), then FDA
   encourages manufacturers to petition for a health claim for
   their particular product. The same science will be applicable,
   regardless of whether the petition is for a single
   supplement-type product, a clearly specific ingredient, or a
   specific type of soluble fiber contained in foods.
   
   III. Comments
   
   A. Food Claims Versus Nutrient Claims
   
       1. In the proposal on general requirements for health
   claims (56 FR 60537), FDA specifically requested comments on
   issues relating to health claims on the label or labeling of
   foods which targeted foods rather than nutrients, criteria to
   identify foods eligible for such claims, and possible measures
   to assure that consumers are not misled as to the benefit of
   consuming specific products.
       A variety of comments submitted in response to the
   proposed rule on dietary fiber/CVD supported claims on foods.
   Specifically, relative to a claim for dietary fiber and heart
   disease, a few comments stated that FDA does have the
   authority to regulateypes of foods, stating that generous
   intakes of such foods in diets may help lower the risk of
   heart disease and certain forms of cancer. Other comments
   stated that, at this time, a scientific basis does not exist
   on which to authorize a specific claim for dietary fiber and
   heart disease, and that a food claim was more appropriate.
   Other comments suggested that such claims be developed in
   concurrence with ``The Surgeon General's Report on Nutrition
   and Health,'' (Surgeon General's Report) (Ref. 63)
   andobjectives identified in ``Healthy People 200: National
   Health Promotion and Disease Prevention Objectives,'' (Healthy
   People 2000) (Ref. 64), and should focus on the total diet.
       A number of comments stated that the amount of fat in the
   American diet should be lowered and the amount of dietary
   fiber increased through a variety of food choices from among
   all the food groups. Several comments favored a limited claim
   at point of purchase for foods that would help increase
   dietary fiber intake and lower fat intake. A number of
   comments noted that an appropriate health statement on food
   labels should emphasize that eating a variety of food sources
   of dietary fiber daily (cereals, grains, fruits, and
   vegetables) can decrease the risk of certain forms of cancer
   and heart disease. A number of comments stated that food
   claims, if allowed, should not give the impression that a
   single nutrient (for example, dietary fiber) is responsible
   for the reduction in risk of heart disease. Several comments
   suggested that a health claim regarding dietary fiber and
   heart disease be made only on fruits, vegetables (fresh or
   processed), or whole grains and not on products that combine
   these foodswith other ingredients (such as bread products or
   sauce). Other comments favored use of claims on foods but not
   supplements.
       One comment stated that FDA should allow a fiber health
   claim because virtually all dietary guidelines for Americans
   have encouraged consumption of fiber-rich foods, including
   whole grain cereals, fruits, and vegetables, and that
   comprehensive government and other reviews by recognized
   scientific bodies concluded that dietary patterns that include
   fiber-rich foods are associated with reduced risk of
   colorectal cancer, heart disease, and other chronic diseases.
       As the agency has discussed, in the final rule on general
   requirements for health claims published elsewhere in this
   issue of the Federal Register, statements about good nutrition
   that do not, expressly or by implication, refer to a substance
   are considered dietary guidance and not health claims. In this
   rule, the agency has concluded that the scientific evidence is
   sufficient to support a health claim that refers to a
   substance contained in certain fruits, vegetables, and grain
   products and relates those foods to a reduced risk of heart
   disease. Specific reference to the fact that these foods
   contain soluble fiber is authorized, since this nutrient
   serves as a useful marker for the broad product categories of
   foods which correlate with reduction in blood LDL-cholesterol
   levels, and consequently, with reduced risk of heart disease.
   Thus, the agency has been persuaded by the comments that the
   totality of the evidence supports a health claim which
   identifies foods whose use is protective against heart disease
   and whose selection can be facilitated by reference to the
   marker nutrient, soluble fiber. Because soluble fiber is
   usually considered a useful adjunct to, but not a replacement
   for, a diet low in saturated fat and cholesterol (Ref. 66),
   the agency is also requiring this information in the label
   claim.
   
   B. General Comments
   
       2. Several comments supported a health claim for dietary
   fiber and heart disease, stating that there is sufficient
   scientific evidence to support such a claim or that a claim is
   warranted because heart disease and, hence, CVD are major
   public health p consume high fiber diets have a low incidence
   of heart disease, these claims should be allowed. Several
   comments stated that FDA should consider the rapid pace of
   advances in knowledge that link nutritional substances to good
   health and disease prevention. Other comments stated that a
   claim regarding dietary fiber and CVD should not be allowed
   because overall health depends on a number of factors, such as
   exercise and lifestyle characteristics.
       FDA agrees that CVD and, consequently, CHD are significant
   public health problems. The agency, in the proposal,
   tentatively concluded that diets high in fiber-rich foods,
   including whole grains, fruits and vegetables, are associated
   with reduced risk of CHD, and thus CVD. In the proposal, the
   agency also noted that these diets differ in levels of many
   nutrients, such as saturated fat and vegetable protein, and in
   types of dietary fiber, making it difficult to ascribe, from
   observational studies on whole diets, the observed nutrient
   and disease relationship to a single dietary component (56 FR
   60582 at 60592 and 60593).
       Several new studies that became available after
   publication of the proposal were suggestive of positive
   effects of soluble fiber intakes on blood total- and
   LDL-cholesterol levels, risk factors for heart disease.
   However, FDA has also concluded, as noted in the proposal,
   that the effectiveness of these fibers may be affected by
   other dietary components (e.g., the level of saturated fat and
   cholesterol in the diets), as well as by physical
   characteristics (e.g., particle size or water-holding
   capacity), or by the fiber source itself. Thus, while the
   agency has concluded that not all soluble fibers, i.e., as
   identified by the AOAC method for soluble fiber determination,
   are effective in lowering cholesterol, and other components of
   fiber-rich foods,i.e., vegetable proteins or lipids, may
   contribute to the cholesterol-lowering effect observed. In
   addition, the hypocholesterolemic effectiveness of some
   soluble fibers has been reported in studies in which the
   source of soluble fiber was consumed as an adjunct to a low
   saturated fat, low cholesterol, and low total fat diet.
       3. Several comments stated that FDA is not following the
   congressional mandate to consider whether there is significant
   scientific agreement supporting specific health claims.
   Specifically, the comments argued that the agency should have
   placed its inquiry in the proper context by first identifying
   the range of specific health claims that could be made about
   dietary fiber and CVD, and then examining the scientific
   support for each claim.
       FDA disagrees with this comment. The 1990 amendments did
   not require the agency to identify the wide range of health
   claims that might be made with respect to the 10 topics
   identified in the act and then to evaluate all published
   studies relevant tothe claims. Rather, the 1990 amendments
   instructed the agency to determine whether claims respecting
   the 10 areas, including ``dietary fiber and cardiovascular
   disease'' meet the requirements of section 403(r)(3) or
   (r)(5)(D) of the act. The agency interpreted this directive in
   a straightforward and logical way. Indeed, FDA's chosen
   approach was necessary if the agency hoped to accomplish the
   congressional mandate within the required timeframe. Thus,
   FDA, in its proposed rule (56 FR 60582), focused itsreview of
   the science on those aspects of the dietary fiber and CVD
   relationship for which the strongest scientific evidence
   exists: Soluble fiber and CHD.
       4. Some comments stated that FDA's denial of a health
   claim for dietary fiber and CVD, because of rigid application
   of a scientific standard higher than that mandated by the 1990
   amendments, would have unfortunate public health consequences.
       FDA disagrees with the comment that the agency is applying
   a standard higher than that mandated by the 1990 amendments.
   To ensure the validity of health claims, Congress enacted a
   scientific standard in section 403(r)(3)(B)(i) of the act. FDA
   intends specifically, any claim for which, based on the
   totality of the publicly available scientific evidence, there
   is significant scientific agreement, among experts qualified
   by training and experience to evaluate such claims, that the
   claim is supported by the evidence. FDA also disagrees that
   applying the scientific standard mandated in section
   403(r)(3)(B)(i) of the act will have unfortunate public health
   consequences. FDA believes that for health claims to be truly
   educational and provide public health benefits, they must be
   scientifically valid and not misleading. The issue of the
   scientific standard is discussed in more detail in the final
   rule on general requirements for health claims, published
   elsewhere in this issue of the Federal Register.
       5. Some comments stated that FDA used disparate criteria
   in assessing the relationship between lipids and CVD and
   dietary fiber and CVD, but did not elaborate on this issue.
       FDA disagrees with these comments. In reviewing the
   scientific literature for the development of its proposed
   rules for health claims, FDA followed the standard mandated by
   the 1990 amendments.
       Federal Government reports and other authoritative
   documents have consistently concluded that there is a strong
   relationship between the total amount and types of dietary fat
   and other lipids in the diet and the risk of heart disease. In
   developing the proposed rule on lipids and CVD, the agency
   found that new evidence supported these conclusions. The
   weight of the evidence showed that diets low in saturated fat
   and cholesterol are associated with reduced blood total- and
   LDL-cholesterol and a lower risk of CHD.
       In contrast, Federal Government reports and other
   authoritative documents did not reach similar conclusions that
   the scientific evidence supported a claim that dietary fiber
   per se is associated with the reduced risks of CVD. The
   available evidence showed an association between consumption
   of diets high in fruits, vegetables, and grain products--diets
   which are generally high in fiber--and risk of heart disease.
   For example, in its recommendations in the NAS report ``Diet
   and Health'' (Ref. 48), the committee on Diet and Health
   ``agree[d] with most other expert groups in proposing that the
   intake of vegetables, fruits, and other sources of complex
   carbohydrates should be increased and that the intake of
   sugars should be limited.'' The committee further noted that
   ``the strength of the evidence does not justify making
   specific recommendations pertaining to dietary fiber at this
   time. The committee's recommendation to emphasize the
   consumption of vegetables, fruits, and other sources of
   complex carbohydrates would, however, indirectly result in
   increased consumption of dietary fiber.''
       In developing its proposed rule on dietary fiber and CVD,
   the agency found that the evidence available at the time the
   proposal was developed did not alter these conclusions. The
   agency found that the scientific evidence was not sufficiently
   conclusive or specific for dietary fiber per se versus other
   components in the diet to justify use of a health claim
   relating intake of dietary fiber to reduced risk of CVD.
       6. One comment stated that FDA failed to comply with the
   1990 amendments (section 403(r)(4)(C) of the act) in that it
   has rejected the conclusion of authoritative Federal
   Government reports without justifying its decision to do so as
   the act requires. The comment stated that the National
   Cholesterol Education Program (NCEP) has concluded that
   soluble fiber may help reduce blood cholesterol levels. The
   comment refers to the NCEP 1989 consumer pamphlet (Ref. 5),
   which recommends breads, pasta, rice, cereals, dried peas and
   beans, fruits, and vegetables as good sources of complex
   carbohydrates (starch and fiber). The comment quotes from the
   pamphlet that these foods are ``excellent substitutes for
   foods that are high in saturated fat and cholesterol. Teduce
   blood cholesterol levels'' (Ref. 5). The comment also noted
   that the NCEP expert panel report, ``Population Strategies for
   Blood Cholesterol Reduction'' (Ref. 66), supports the
   recommendation to consume vegetables, fruits, breads, legumes,
   and whole grain cereals. The comment quotes the NCEP report
   that ``Dietary fiber supplements are not a panacea for blood
   cholesterol problems. Foods rich in soluble dietary fiber are,
   however, a useful addition to a low saturated fatty acid, low
   fat, and low cholesterol eating patterns* * *'' and:
       ``Oat bran exhibits hypocholesterolemic properties due to
   its appreciable content of oat gum. Soluble fibers such as
   pectin, guar gum, locust bean gum, or psyllium in large
   quantity supplementation have been shown to lower total and
   LDL-cholesterol levels. The absolute effect on LDL-cholesterol
   concentrations is modest even when the amount of soluble fiber
   such as oat bran is consumed in appreciable amounts (60 g).
   This effect, however, represents a useful adjunct to an eating
   pattern low in saturated fatty acids and cholesterol'' (Ref.
   65).
       FDA disagrees that, in developing its proposed rule
   regarding fiber and CVD, it rejected conclusions of Federal
   Government reports. Comments, through repetition of those
   portions of the text that accompanies dietary recommendations
   and that includesthe words ``soluble fiber,'' are attributing
   greater significance to the statements relating soluble fiber
   to heart disease risk than was given to these results by the
   expert panels. This selected emphasis distorts the meaning of
   the authoritative reports in question by failing to
   acknowledge important contributions to reduced risk of disease
   by the wide variety of nutrients and nonnutritive substances
   present in diets high in fruits, vegetables, and grain
   products. Such an emphasis also focuses attention away from
   changes in overall dietary patterns and their potential
   contributions to reducing risk of chronic diseases.
       In the NCEP report (Ref. 65) cited by the comment, the
   expert panel noted the hypocholesterolemic effects of some
   soluble fibers, but recommends ``a habitual pattern of eating
   that is consistently low in saturated fatty acids, total fat,
   and cholesterol.'' NCEP further recommended that ``all healthy
   Americans recognize that no single food or supplement is the
   answer to achieving a desirable blood cholesterol level''
   (Ref. 65). NCEP's recommendation to Americans is to ``eat a
   greater quantity and variety of fruits, vegetables, breads,
   cereals, and legumes'' (Ref. 65). These food choices ``will
   help to meet nutritional needs for minerals, vitamins, dietary
   fiber (including soluble fiber), and complex carbohydrates,
   and to replace calories from fat.'' Thus, the NCEP
   acknowledges the importance of a dietary pattern that focuses
   on reducing fats in the diet in order to reduce serum
   cholesterol. It did not attribute a protective effect from CVD
   to dietary fiber alone.
       Neither the Surgeon General's report on ``Nutrition and
   Health'' (Ref. 63), the National Academy of Sciences (NAS)
   ``Diet and Health: Implications for Reducing Chrinic Disease
   Risk''(Diet and Health) (Ref. 48), nor (The Department of
   Health and Human Services (DHHS)) ``Healthy People 2000''
   (Ref. 64) found the scientific evidence strong enough to
   attribute the protective effects against CVD of dietary
   patterns high in fruits, vegetables, and grain products
   exclusively to the soluble fiber content of such diets.
   Specifically, the Surgeon General's report on ``Nutrition and
   Health'' (Ref. 63) recommends increased consumption of whole
   grain foods and cereal products, vegetables (including dried
   beans and peas) and fruits. The report states that:
       ``the association shown in epidemiologic and animal
   studies between diets high in complex carbohydrates and
   reduced risk for CHD and diabetes mellitus is, however,
   difficult to interpret. The fact that such diets tend also to
   be lower in energy and fal studies also suggest that
   water-soluble fibers from foods such as oat bran, beans, and
   certain fruits are associated with lower blood glucose and
   blood lipid levels'' (Ref. 63).
   The section concludes with the statement, ``Current evidence
   suggests the prudence of increasing consumption of whole grain
   foods and cereals, vegetables (including dried beans and
   peas), and fruits'' (Ref. 63).
       Similarly, the Executive Summary of the National Research
   Council's ``Diet and Health'' recommends, ``Every day eat five
   or more servings of a combination of vegetables and fruits,
   especially green and yellow vegetables, and citrus fruits.
   Also, increase intake of starches and other complex
   carbohydrates by eating six or more daily servings of a
   combination of breads, cereals, and legumes'' (Ref. 48). The
   summary continues ``Studies in various parts of the world
   indicate that people who habitually consume a diet high in
   plant foods have low risks of atherosclerotic CVD's, probably
   largely because such diets are usually low in animal fat and
   cholesterol. Some constituents of plant foods, e.g., soluble
   fiber and vegetable protein, may also contribute--to a lesser
   extent--to the lower risk of atherosclerotic cardiovascular
   diseases.'' The Committee does not recommend the use of fiber
   supplements.
       ``Healthy People 2000'' states that recommendations from
   the National Cancer Institute, the Surgeon General's report on
   ``Nutrition and Health,'' NAS' ``Diet and Health,'' and
   ``Dietary Guidelines for Americans'' support increased
   consumption of vegetables, fruits, and whole-grain breads and
   cereals (Ref. 64). ``Healthy People 2000'' also states that
   further research is needed to clarify whether the effect on
   blood lipids is an independent effect, and if so, to quantify
   the relationship (Ref. 64).
       Therefore, in its proposed decision not to authorize the
   use on the label or labeling of foods of health claims
   relating intake of dietary fiber to decreased risk of CVD, the
   agency's tentative conclusion was consistent with those of
   Federal Government and other authoritative reports.
       FDA's position was also consistent with recommendations in
   the Institute of Medicine's report ``Nutrition Labeling:
   Issues and Directions for the 1990s'' (Ref. 81). In this
   report, the authors note that:
       ``there has been a great deal of interest in the specific
   effects of dietary fiber on several chronic diseases. The
   strongest argument for an increase in consumption of dietary
   fiber is the important contribution it makes to normal bowel
   function. Clear scientific associations of fiber intake with
   the incidence of heart disease, certain types of cancer, and
   diabetes mellitus have not been made. One reason may be the
   difficulty in designing appropriate experiments to
   specifically test for the effectof dietary fiber. Foods high
   in dietary fiber are also generally low in calories and total
   and saturated fatty acids and devoid of cholesterol; thus,
   determination of a specific fiber effect in a feeding study is
   difficult. Moreover, foods have a variety of fiber components
   and each may have different actions. Chemically and
   physiologically, cellulose, lignin, hemicellulose, pectin and
   alginate (all relatively purified fiber types) behave
   differently. Wheat bran, oat bran, and rice bran (all
   heterogeneous mixtures of fibers) are not similar in
   composition. It is also very difficult to analyze dietary
   fiber chemically, and thus it is hard to correlate the role of
   specific fiber components to health effects * * *'' (Ref. 81).
       The Institute of Medicine's report (Ref. 81) also provided
   specific recommendations, including: (1) ``FDA and USDA should
   require the disclosure of fiber content per serving in grams
   on the nutrition information panel under the term total
   dietary fiber''; and (2) ``FDA and USDA should discourage
   labeling of soluble or insoluble fiber contents until
   methodologies approved by the agencies allow for the adequate
   and reproducible quantification of the soluble and insoluble
   fiber contents of a variety of
       Therefore, FDA is not rejecting the conclusions of these
   government reports. In its final rule, 101.76, the agency is
   permitting a claim relating dietary consumption of fruits,
   vegetables, and grain products, i.e., good sources of fiber,
   with reduced risk of heart disease.
       7. One comment criticized FDA for misinterpreting the
   conclusions of the Government of Canada's 1985 ``Report of the
   Expert Advisory Committee on Dietary Fiber'' to the Department
   of Health and Welfare (Ref. 46). The comment stated that the
   committee expressly advocated health claims for soluble fiber
   and CHD.
       FDA agrees that the Canadian report in question noted
   hypocholesterolemic effects of some soluble fiber sources, but
   FDA disagrees that the Department of Health and Welfare
   supports health claims on soluble fiber in Canada. In its
   comments to FDA regarding the proposed rule, dietary fiber and
   CVD, the Canadian Department of Health and Welfare stated its
   opposition to health messages. Its comment stated that health
   messages and claims for heart disease and cancer, among other
   diseases, are not permitted under the Canadian Food and Drugs
   Act. These diseases are considered to require medical
   diagnosis and treatment under medical supervision, and thus
   products bearing messages about them are regulated as drugs.
   The 1985 report suggested that, if a company desired to use a
   health claim, then that company should present the evidence of
   the product's effectiveness based on acceptable test protocols
   (Ref. 46).
       8. One comment stated that FDA failed to note, in the
   proposed rule, epidemiological studies cited in the NAS'
   ``Diet and Health'' that found an inverse association between
   dietary fiber and CHD, even after adjusting for the possible
   confounding effects of calories and fat (Ref. 48).
       FDA disagrees with this comment. One study was cited in
   ``Diet and Health'' that showed a protective effect of dietary
   fiber from cereals on risk of CHD independent of caloric
   intake. Results of a study by Khaw and Barrett-Connor (Ref.
   31) (reviewedby FDA in the proposal) found an inverse
   association between fiber intake and ischemic heart disease
   mortality independent of calories among other dietary
   components. However, ``Diet and Health'' also states that the
   authors used 24-hour dietary recall to assess intake, a method
   which has limited applicability in the assessment of the usual
   dietary intake of individuals in the United States (Ref. 48).
   Results from the 1982 epidemiological study by Burr (Ref. 82),
   also cited in ``Diet and Health,'' showed a lower risk of CHD
   in 10,943 vegetarians than in nonvegetarians, but their
   decreased risk could not be accounted for by increased fiber
   consumption, because many other components of the diet also
   differed between these two groups. Therefore, FDA did not
   misinterpret the conclusions of ``Diet and Health'' that there
   is no conclusive evidence that it is dietary fiber, rather
   than other components of fiber-rich foods, that reduces risk
   of CVD.
       9. One comment disagreed with the indication in the
   proposed rule that a ``risk/benefit'' argument was not a
   sufficient or appropriate basis on which to authorize a health
   claim for food labels. The comment stated that, although the
   data for dietaryfiber do not support the hypothesis for
   reduced risk of CVD, the ``potential benefits far outweigh the
   potential disadvantages.''
       FDA disagrees with this comment. Congress enacted a
   scientific standard for health claims in section
   403(r)(3)(B)(i) of the act. Claims must meet the statutory
   requirements; that is, based on the totality of the scientific
   evidence, there must be significant scientific agreement,
   among experts qualified by training and experience to evaluate
   such claims, that the claim is supported by such evidence. The
   concept of ``potential benefits outweighing potential
   disadvantages'' is not an acceptable substit
   
   C. Specificity of the Relationship Between Soluble Dietary
   Fiber and Heart Disease
   
       10. Several comments stated that FDA's refusal to
   authorize a health claim on dietary fiber and CVD is based on
   the agency's determination to treat all dietary fiber as a
   group, rather than considering each fiber source individually.
   The comment stated that dietary fiber is composed of a diverse
   group of materials, as the agency observed, and each has its
   own physiological effects. The comment noted that certain
   water soluble fibers have been documented to reduce serum
   cholesterol, thereby loweringthe risk of CHD.
       FDA disagrees with this comment. In the proposed rule (56
   FR 60582), FDA limited its review of the science to those
   aspects of the dietary fiber and CVD relationship for which
   the strongest scientific evidence exists: soluble dietary
   fiber and CHD. FDA also noted, however, that soluble fiber was
   a heterogenous family of fibers which vary in both chemical
   and physical characteristics. After reviewing the totality of
   the evidence, the agency is persuaded that even if soluble
   fiber alone is effective in reducing risk of heart disease,
   greater specificity than that identified by existing
   analytical methods is needed in order to predict the
   effectiveness of soluble fiber in foods.
       11. LSRO submitted its document, ``Dietary Fiber and
   Cardiovascular Disease'' (Ref. 40), as a comment. In this
   final report, LSRO stated that it remains to be determined
   whether the observed effects of dietary fiber on serum
   cholesterol reduction result strictly from the fiber or from
   other components of the fiber-rich food or from a combination
   of these factors. The report stated that studies suggest that
   soluble fiber, a specific type of dietary fiber, is
   hypocholesterolemic, while insoluble fiber is not. Further,
   when foods are used, foods rich in [-glucans seem to have a
   more hypocholesterolemic effect. The report states that there
   is no indication of optimum level or even a dose-related
   effect, and notes that there are suggestions as to optimum
   level of intake for ``better health'' (e.g., normal bowel
   function) but not for prevention of disease. In addition,
   there are no data relating to transience of fiber effects,
   although this is amenable to experimental testing. The LSRO
   report alsonoted that generalization to the U.S. population is
   difficult. Presumably, persons at high risk, such as those
   with a family history of hyperlipidemia or heart disease,
   would benefit most.
       The LSRO report also states that it is unclear whether the
   lipid-lowering effects observed in some studies are the result
   of the fact that most high fiber diets are low in fat.
   According to the report, most of the available evidence
   suggests that isolated polysaccharides, such as pectin, guar
   gum, locust bean gum, oat gum, and psyllium mucilloid, have
   the ability to lower serum cholesterol levels; however, there
   are no data to indicate that a fiber present in a food is the
   same as when it has been extracted and purified. The data
   suggest that diets high in fiber-rich foods can influence
   lipidemia, but this effect is probably due to overall changes
   in the diet caused by the addition of fiber sources rather
   than simply to a direct effect of fiber.
       One comment stated that FDA failed to cite LSRO's (Ref.
   39) conclusions regarding hypocholesterolemic effects of some
   soluble fibers.
       FDA agrees that the results of clinical studies suggest
   that soluble fiber is hypocholesterolemic, while insoluble
   fiber is not. FDA also agrees that the effect of fiber-rich
   foods on serum lipids is related to the total diet, i.e., one
   that is low in saturated fat and cholesterol and high in
   soluble fiber-rich foods, such as vegetables, fruits, and
   grain products.
       FDA disagrees that the agency failed to consider the
   conclusions of this report. Both of LSRO's reports (Refs. 39
   and 40) concluded that soluble fibers were related to reduced
   blood cholesterol levels, but the LSRO report (Ref. 40) also
   concluded, asts of fiber-rich foods or from a combination of
   these factors. Data available since LSRO's report (Ref. 40)
   provided some additional information as to the effect of
   soluble fiber on blood cholesterol reduction.
       12. One comment stated that FDA failed to note part of the
   World Health Organization's (WHO) report (Ref. 71) on the
   relationship between soluble fiber and blood cholesterol
   levels (and hence CHD). The comment quotes the report,
   ``[o]ther dietary components, such as dietary fibre, have an
   effect on serum cholesterol in experimental studies and are
   correlated in intercountry comparisons. As with fatty acids,
   the different forms of dietary fibre may have different
   effects on serum cholesterol'' (Ref.71). The comment concludes
   that, although the WHO report did not analyze the relevant
   science, it acknowledges the evidence that soluble fibers have
   hypocholesterolemic effects.
       FDA disagrees with the comment that it failed to note the
   WHO statements regarding soluble fiber. Although the WHO
   report states that ``different forms of dietary fiber may have
   different effects on serum cholesterol,'' it does not identify
   which form of dietary fiber affects serum cholesterol; the
   term ``soluble fiber'' was not used in the report (Ref. 71).
       13. Several national health organizations with expertise
   in heart disease agreed with FDA's proposed conclusion, that,
   at this time, there is insufficient evidence to link dietary
   fiber, per se, to CVD. The comments stated that the proposal
   is consistent with the conclusions of all previous expert
   groups. The comments stated that the association between fiber
   and blood lipids is specific to soluble fiber and that
   specificity to fiber class is unresolved. The amount of
   soluble fiber necessary to produce blood cholesterol lowering
   is unclear; nor is it known whether (and if so, how much) the
   response differs by type of soluble fiber (i.e., [-glucan
   versus pectin). Furthermore, the comments state that FDA's
   review of the scientific literature (56 FR 60574) mentions the
   tumor-enhancing effect of soluble fibers in animal studies.
   They recommended that FDA not allow health claims that link
   fiber to risk of CVD's. The comments stated that there are
   insufficient data to warrant such a claim and thatit is
   misleading to permit a claim that singles out a particular
   food or foods in diets. The comments stated that if there were
   sufficient data to make a claim, it should be stated in the
   context of a low fat, low cholesterol diet. It would also be
   necessary to specify the type of fiber, e.g., soluble fiber,
   in the case of CVD.
       FDA agrees that a health claim for a dietary component
   should be stated in the context of the total daily diet and,
   in this case, should be specific to the type of dietary fiber.
   The agency has been persuaded, based on its review of the
   comments andits review of the scientific literature, that
   questions remains as to whether the cholesterol-lowering
   effect observed with some soluble-fiber food sources (e.g.,
   oats) is due to the soluble fiber component or to a
   combination of other components associated with these foods.
       14. One comment questioned the motivation behind the
   agency's tentative rejection of health claims for fiber and
   CVD. The comment stated that FDA does not want any health
   claims on dietary supplements and that the agency should not
   preclude claims because of concern that dietary supplement
   manufacturers will then be able to make such claims. Comments
   from supplement manufacturers stated that, if health claims
   are permitted on fiber-containing foods, then fiber
   supplements should also be permitted to carry a claim because
   there is no difference between fiber in foods and fiber in
   supplements and all fiber supplements are safe. Other comments
   stated that FDA should authorize health claims on supplements
   because supplements offer an alternative to cons
       FDA disagrees with the comment that it does not want
   health claims on fiber supplements and with the suggestion
   that it rejected dietary fiber and CVD health claims because
   of concern that supplement manufacturers would then be allowed
   to make such claims. FDA has applied the law equally to
   supplements and conventional foods.
       As the agency's proposed rule stated, FDA proposed to deny
   a health claim for dietary fiber and CVD because the agency
   tentatively concluded that the available scientific evidence
   was not sufficiently conclusive or specific for dietary fiber
   per se.FDA notes that the comment claimed that there is no
   difference between fiber in foods and fiber in supplements but
   submitted no data were submitted to support this statement.
   Indeed, several expert reports (Refs. 39, 40, and 46)
   concluded that there is no evidence that fiber, when isolated
   and/or processed in foods, has the same physiological effects
   on serum cholesterol as consumption of the native fiber from
   fiber-rich foods. These reports note that the predictive
   capability of analytically determined values for soluble fiber
   and physiologic activity has not been established, and
   effectiveness may vary by source of fiber or by physical
   characteristics not detected with chemical methods of analysis
   (e.g., particle size or water-holding capacity). They also
   note that safety may vary between native and isolated sources
   of fiber.
   
   D. Comments Regarding a Relationship Between Specific Soluble
   Dietary Fibers and CVD
   
       15. One comment submitted a review of available literature
   for particular fibers. The comment stated that several of
   these fibers are ``effective as cholesterol lowering agents
   and in addition, they are safe to use provided a few
   reasonable precautions are taken.'' The comment identified
   fibers such as locust bean gum, guar gum, oats, pectin, and
   psyllium mucilage as materials with hypocholesterolemic
   effects. Less well-tested fibers that have some
   hypocholesterolemic effects (for example, barley, acacia gum,
   dried beans, and karaya gum) were also mentioned. The comment
   stated that wheat fiber requires more study to determine its
   effects. The comment notes that, since dietary fiber
   represents a diverse group of materials, FDA should consider
   allowing health claims and statements on individual dietary
   fiber materials. The comment also recommended separate health
   claims on fibers with ``hypocholesterolemic'' activity,
   because each of these fiber materials also has a different
   dose-response. The comment provided some criteria on which to
   base the individual fiber claims.
       FDA agrees that the scientific literature shows that
   dietary fiber is a complex group of dietary substances with
   differing chemical, physical, and physiological properties,
   and that not all soluble fibers are alike in their
   hypocholesterolemic properties. FDA disagrees with the
   recommendation to allow individual health claims on various
   soluble fibers, but does concur that the effectiveness of
   individual fibers in foods may be documented for specific food
   products or for fibers whose physical and chemical
   characteristics are well specified (e.g., oat brans meeting
   specified parameters). Thus, if manufacturers can document,
   through appropriate studies, that the soluble fiber in their
   particular food is effective in lowering LDL-cholesterol, and
   hasno adverse effects on other heart disease risk factors
   (e.g., HDL-cholesterol), then FDA encourages manufacturers to
   petition for a health claim for their particular product or
   ingredient. Additionally, if new evidence becomes available
   which more clearly identifies what type of soluble fiber is
   effective, this also would be appropriate for a petition.
       16. Another comment stated that part of FDA's difficulty
   in interpreting data on dietary fiber and CVD results from the
   consideration of soluble fiber as a single nutrient instead of
   as a class of diverse substances. The comment stated that data
   existerance of literature supporting oat bran and its
   hypocholesterolemic effects and that the active mechanism
   behind oat bran's effectiveness is the [-glucan component. The
   comment suggested that the agency allow foods to be identified
   that contributeto eating patterns that reduce the risk of
   disease.
       FDA agrees, as discussed in the proposal, that dietary
   fiber is a diverse group of substances and not all soluble
   fibers are alike in their hypocholesterolemic properties. FDA
   also agrees, based on new data submitted as comments, that
   there is evidence to suggest that [-glucan has
   hypocholesterolemic properties. However, as noted in the
   meta-analysis (Ref. 125), other components of oats, such as
   tocotrienols and oat oil, may play a role in the reduction of
   blood LDL-cholesterol levels. However, FDA is authorizing a
   claim on fruits, vegetables, and grain products--foods that
   are good sources of soluble fiber and other substances that
   may have cholesterol-lowering properties.
       17. Comments from a health professional organization
   stated that several studies have shown that soluble fibers,
   such as those in oat, beans, psyllium, and guar, appeared to
   lower serum cholesterol, at least when given in large amounts.
   The comment further stated that weight loss observed in some
   studies may possibly be far more important than fiber in
   contributing to cholesterol reduction.
       FDA agrees that weight loss associated with test diets may
   affect the ability to differentiate between the effects of the
   test substance (i.e., soluble fiber) versus the
   well-documented effect of weight loss on blood total- and LDL-
   cholesterol. This was discussed in the meta-analysis submitted
   as a comment (Ref. 125). Several studies observed reductions
   in blood cholesterol levels in studies in which loss did not
   occur. For example, in a study by Marlett et al. (Ref. 124),
   submitted with another comment, there was no significant
   weight loss in either the control or soluble fiber (oat bran)
   groups during the fiber intervention period. Serum cholesterol
   was significantly lower in the oat bran group than in the
   control group. Spiller et al. (Ref. 87) reported no weight
   changes in two groups receiving either guar gum or oat fiber.
   Both groups experienced significantly lowered serum
   cholesterol. In two other studies (Refs. 118 and 120), there
   was no significant difference in the weight loss between
   subjects consuming oat bran (Ref. 120) or psyllium cereal
   (Ref. 118) and those consuming wheat bran. The oat bran group
   had significantly lower serum cholesterol than the wheat group
   and compared to the control period. Thus, while the agency
   agrees that weight loss may be a confounding factor, the new
   evidence is consistent with the concept that consumption of
   diets high in certain soluble fiber-rich foods, independent of
   body weight, has a beneficial effect on blood cholesterol
   levels.
   
   E. Comments Regarding FDA's Interpretation of Specific Studies
   
       Some comments cited references that were already reviewed
   by FDA or the consensus documents, or were studies of
   questionable relevance to human CVD (e.g., studies evaluating
   postprandial glucose response of fiber diets). Some abstracts
   were also cited that do not provide sufficient information for
   evaluation. Review articles that provided no new data were
   also included with some comments.
       18. One comment stated that the study by Kahn et al. (Ref.
   25) showed significant decreases in relevant parameters in the
   test groups versus baseline and that the authors concluded
   that soluble fiber ``appears to be quite effective'' in
   lowering those parameters. In the proposed rule, FDA reported
   that there was no significant difference in serum cholesterol
   between the test group and the control group.
       FDA notes that it did report that a significant difference
   in serum cholesterol existed between baseline and intervention
   values if the results are examined without consideration of
   the immediate versus delay constraint of the design. The
   agency initit of oat bran on serum cholesterol lowering
   because of the lack of statistical significance when examining
   only the difference between the immediate intervention group
   and its control. FDA believes that the study design should
   have been modified to overcome this problem of the time delay
   in comparing the groups. However, after further review, FDA
   agrees with the authors (Ref. 25) that, after correcting for
   the time delay, the study does show that oat bran
   supplementation reduced serum cholesterol.
       19. One comment criticized FDA for calling a ``p value''
   of 0.052 in the paper by Anderson et al. (Ref. 5) ``nearly
   significant.'' The comment suggested that a more precise term
   would be ``borderline significant.''
       FDA acknowledges that the term ``borderline significant''
   is more commonly used professionally, but considers that this
   does not negate the need for caution in reviewing the results
   of this paper.
       20. One comment stated that FDA failed to report the
   modest effect, from a study by Bell et al. (Ref. 12), of a
   pectin-enriched cereal on serum cholesterol.
       FDA disagrees with this comment. Of the two fiber-enriched
   cereals (one enriched with psyllium and the other with pectin)
   tested by Bell et al. (Ref. 12), FDA reported that only
   psyllium demonstrated a significant reduction in cholesterol.
   The agency correctly stated that the psyllium-enriched diet
   demonstrated significant lowering of total- and
   LDL-cholesterol, but that the pectin did not.
       21. A comment disagreed with FDA's statement that the
   study by Davidson et al. (Ref. 15) did not demonstrate
   significant reductions from baseline levels in total
   cholesterol with daily intakes of up to 2 oz of oatmeal (1.2
   to 2.4 g [-glucan/day) or 1 oz oat bran (2.0 g [-glucan/day)
   in persons on a Step 1 diet.
       FDA correctly reported that only the higher fiber intake
   groups showed statistically significant effects. The small
   sample size may have prevented seeing an effect in the lower
   doses. FDA agrees that, based on the results of this study, an
   intake ofsoluble fiber (in this case, [-glucan from oats) of
   about 3 g per day or more was beneficial in that it resulted
   in a significant lowering of serum cholesterol in persons
   consuming a low-fat diet.
       22. A comment asked FDA to consider a result from a study
   by Newman et al. (Ref. 50) which used [-glucan from barley and
   compared it to wheat fiber. The cholesterol levels of the high
   [-glucan group did not differ significantly from baseline
   after four weeks. The author explains this lack of
   significance by citing the small sample size and small amounts
   of [-glucan consumed.
       FDA does not necessarily disagree with this comment, but
   considers that it was correct in not assessing the results as
   positive (i.e., a true difference exists) because a
   statistically significant effect was not demonstrated.
       23. One comment stated that FDA failed to consider
   significant the results by Van Horn et al. (Ref. 68). In this
   study, the investigators tested the effects of oat-bran
   intervention on serum cholesterol levels in subjects with
   normal or mildly elevated serum cholesterol levels. The
   authors reported a difference (p = 0.074) in serum cholesterol
   between the oat group and the control after 8 weeks.
       FDA did not consider this result significant because the
   small sample size was not a problem in this study and the p
   values reported are one-tailed. Such values allow leeway over
   the more conventional and conservative two-tailed tests.
       24. One comment stated that FDA did not properly interpret
   the results of a study by McIntosh et al. (Ref. 44), which
   compared the cholesterol-lowering effect of [-glucan from
   barley to that of wheat fiber.
       FDA disagrees with this comment. Although there was a
   significant difference between cholesterol levels of the two
   groups, this was largely due to a rise in total cholesterol
   that occurred in the wheat group compared to baseline. In this
   case, it is nat itself is cholesterol-raising. However,
   because the group fed barley did not show a significant
   lowering of cholesterol from their own baseline level, the
   agency does not find that a positive effect of barley was
   demonstrated.
       25. Another comment addressed a study by Burr et al. (Ref.
   13) that examined whether there was a change in mortality
   rates between men given dietary advice to increase fiber
   intake or to reduce fat intake. The comment was specifically
   concerned withFDA's conclusion that the study was not able to
   show a difference between mortality rates in the two groups.
   The comment stated that the results show that increasing fiber
   intake is no less effective than decreasing fat intake in the
   prevention of CHD.
       FDA disagrees with this comment. Because there was no
   group in this study which acted as a control (no changes in
   diet), the study does not provide evidence that either diet in
   the studies is effective or that neither is effective. The
   agency finds that the study does not add to the evidence of
   the effects of fiber.
       26. One comment stated that a study by Little et al. (Ref.
   42) that shows that lowering fat is responsible for
   significant lowering of total cholesterol, while increasing
   fiber is not, did not specify the type of fiber (i.e., soluble
   versus insoluble) and should not be considered as part of the
   fiber-CHD evaluation.
       FDA disagrees with this comment. The agency considered any
   fiber study which might be relevant to the health claim.
   However, the weight given to study results was influenced by
   several factors, including the quality and usefulness of the
   informationon soluble fiber intakes.
       27. One comment noted that the study by Demark-Wahnefried
   et al. (Ref. 16) showed a significant drop in total serum
   cholesterol in all three of the following dietary groups: A
   high fiber, low-fat diet; a low-fat diet alone; and a regular
   diet with fiber added. The comment contended that, because
   there was no significant difference between any of the groups,
   this study showed that increasing soluble fiber intake is as
   effective as reducing fat intake.
       FDA disagrees with this comment. This study does not
   support a conclusion that soluble fiber is as effective as
   reducing fat intake. These results showed that the
   cholesterol-lowering effect of a low-fat diet was not further
   enhanced by the additionof fiber. Additionally, there was a
   significantly decreased consumption of fat and cholesterol in
   all groups; this is consistent with the hypothesis that fat
   displacement is one mechanism by which high fiber diets lower
   total cholesterol.
       28. Another comment also criticized FDA's review of the
   Demark-Wahnefried et al. study (Ref. 16). The comment stated
   that, according to the Keys equation for predicting changes in
   blood cholesterol levels from changes in intakes of fatty
   acids, only40 percent of serum cholesterol's lowering is
   explained by changes in dietary fat.
       FDA disagrees with this comment. The agency's evaluation
   of this study pointed out the lack of significant difference
   between cholesterol lowering on the regular diet plus oats or
   the low fat, low cholesterol diets. The study did not estimate
   how much, if any, of the remaining 60 percent is attributable
   to oat bran.
       29. One comment stated that FDA misinterpreted the results
   of the study by Keenan et al. (Ref. 26) by failing to note
   that the control group decreased its soluble fiber intake over
   the study period, thus accounting for the return to baseline
   for serum cholesterol in this group by the end of the study.
       FDA disagrees that it misinterpreted the results of this
   cross-over study, in which the authors compared the
   hypocholesterolemic effects of oat- or wheat-supplemented Step
   1 diets with a control group on the Step 1 diet only. Although
   FDA recognizes tr over the study period may suggest redoing
   the study with a better control of soluble fiber in this
   group.
       30. Another comment also disagreed with FDA's criticism
   that the study by Keenan et al. (Ref. 26) did not have a
   placebo. The comment stated that the wheat group was
   considered a placebo because it has been well documented that
   wheat has virtually no cholesterol lowering properties. A
   similar comment was made regarding FDA's criticism of a study
   by Swain et al. (Ref. 57) in which wheat was also used as a
   placebo.
       FDA has expressed concern about the use of wheat as a
   placebo due to its inconsistent effect on serum cholesterol in
   some reports. Although the study by Keenan demonstrated an
   increased total cholesterol in the ``oat-wheat'' group of 6
   percent abovebaseline during the wheat period, in Swain's
   study the wheat group showed a 7 percent decrease in serum
   cholesterol. Based on these studies and other reviews, FDA now
   believes that these variations in serum cholesterol may be a
   function of the amount and method of administration of the
   wheat and must be reviewed with caution. Rather than
   discounting the use of wheat as a placebo, FDA believes that
   the placebo, whether it is wheat or another fiber, must be
   evaluated individually for each study.
       31. One comment suggested that the FDA failed to note that
   the study by Davidson et al. (Ref. 15) was specifically
   designed to determine whether [-glucan has a dose-response
   effect on serum cholesterol after reducing and controlling for
   fat intake. Subjects in this study consumed a Step 1 diet with
   two levels of oat bran incorporated into the diet. FDA
   disagrees that it failed to note this point. The authors of
   the study stated that there is a ``lack of continued dose
   response'' (Ref. 15). In addition, the authors say that ``the
   impact of fat substitution on serum cholesterol reduction with
   oat cereals cannot be completely excluded'' due to the lack of
   isocaloric control for the higher-dose treatment groups (Ref.
   15).
       32. Another comment on the Davidson et al. study (Ref. 15)
   stated that FDA's criticism of the lack of a control group was
   not valid due to the 6-week wash-out period. The comment
   states that, during this wash-out period, the serum
   cholesterol valuesof all treatment groups returned to baseline
   levels, suggesting that the hypocholesterolemic effects
   observed were the result of oat [-glucan supplementation.
       Although the wash-out period is not the equivalent to a
   true control, FDA agrees that, considered with the evidence
   from the intervention period, the decrease in serum
   cholesterol suggests an effect of [-glucan at the higher
   levels of intake.
       33. One comment criticized FDA for failure to note
   regarding the study by Bell et al. (Ref. 11) that the psyllium
   group on the Step 1 diet had significant reductions in total
   cholesterol levels as compared to the placebo (wheat fiber)
   controls.
       FDA disagrees with this comment. The problem with this
   study is that, during the second 8-week test period, the
   psyllium group's total cholesterol had risen to only 1.5
   percent below the baseline level. The significant difference
   still exists between placebo and test group because the mean
   serum cholesterol level of the placebo group actually
   increased to a level higher than baseline. The agency believes
   that these results weaken the support for cholesterol-lowering
   effectiveness over time on this diet.
       34. One comment stated that the study by McIvor et al.
   (Ref. 45), showing no effect of guar gum on serum cholesterol
   in a population of overweight noninsulin dependent diabetics,
   should not have been discussed in the proposed rule. The
   comment explained that this was not a typical population group
   (diabetics consume large amounts of fiber products to help
   with carbohydrate metabolism), and the objective of the study
   was to test the safety of guar. In addition, the comment
   states that the diets were urporated.
       FDA disagrees with this comment. Although overweight
   diabetics may not be a typical group, FDA did not eliminate
   the noninsulin diabetics from the criteria for evaluating
   fiber studies; insulin-dependent diabetics are excluded,
   however. Although theobjective of the study was not to
   evaluate cholesterol-lowering effects, this information was
   presented and FDA considered it to be relevant. In addition,
   adding fiber to a diet without changing the overall diet is
   probably a very typical form of consumer behavior. The agency
   also notes that it did not say that fiber intake had no effect
   in this study, but rather that the results are
   ``inconclusive.''
       35. One comment addressed FDA's review of the study by Lo
   and Cole (Ref. 43), in which pooled periods only show a
   significant difference between the placebo and soy fiber
   groups. The comment stated that the placebo's effects on total
   cholesterol were insignificant, while soy's effects were
   significant.
       FDA disagrees with this comment. As shown in table IV of
   the paper, the 2 percent decrease in serum cholesterol
   reported after soy fiber consumption in Group B is not
   significant. FDA questions the inconsistency of the results
   (i.e., after the placebo effect is gone, is Group B more
   representative of the effect?). The order of treatment effects
   on the results should have been considered in the conclusion.
       36. One comment criticized FDA's review of the study by
   Superko et al. (Ref. 56), which FDA discounted as
   demonstrating significant differences between fiber groups.
   The comment stated that FDA has failed to consider the
   significant reductions in serum cholesterol levels between
   baseline and test group consuming guar gum.
       FDA disagrees with this comment. While it is true that at
   4 weeks there was a significant difference between total
   cholesterol levels from baseline, at 8 weeks the p value
   dropped to 0.15, which is not statistically significant. When
   FDA calculated the difference between the placebo to guar
   supplementation at the 4 and 8 week intervals, neither was
   significant as defined by a p value of 0.05 or less.
       37. One comment stated that FDA incorrectly reported
   results in the study by Beling et al. (Ref. 10). The comment
   stated that FDA reported nonsignificant reductions in serum
   cholesterol between the group using oat-enriched cereal and
   the control group.
       FDA agrees with this comment. After reexamining the
   results, FDA notes that they show a significant difference
   between the two groups, thus adding support to a
   cholesterol-lowering effect of soluble fiber-rich foods in
   combination with consumption of a low-fat diet.
       38. Another comment stated that the weight loss
   differences reported in the study by Beling et al. (Ref. 10)
   were not significant between groups, so effects of weight
   change should be similar. The comment also stated that it is
   not possible to completely ``blind'' a study that uses
   ready-to-eat cereal.
       FDA agrees with the comment.
       39. One comment stated that FDA criticized the lack of
   baseline fiber intake data in the study by Gold and Davidson
   (Ref. 19), but did not explain the significance of that
   assertion.
       FDA considers baseline fiber intake data necessary to
   determine if the effect seen could have resulted from fat
   displacement in the diet.
       40. One comment criticized FDA's evaluation of a study by
   Stewart et al. (Ref. 55) and stated that FDA failed to
   consider the effect after adjustment when reporting the study
   as nonconclusive of an effect. The study by Stewart and
   colleagues (Ref. 55) noted that the psyllium supplement
   administered had a nonsignificant effect on total serum
   cholesterol when dose was not considered, but a significant
   effect after adjusting for dose.
       FDA disagrees with this comment. Referring to both Figure
   1 and Table 4 in the study, the linear trend is driven by the
   higher dose values, which have very small numbers compared to
   the overall study population. In addition, the author has
   excluded thr a dose of 0 (no psyllium intake), which, if
   included, would most likely eliminate any trend. The results
   of the study do not support any overall effect of psyllium on
   serum cholesterol.
       41. One comment stated that FDA unnecessarily
   distinguishes studies using separate fiber supplements from
   those using fiber-enriched foods.
       FDA disagrees with this comment. FDA is applying a
   consistent legal standard to its consideration of studies on
   fiber supplements and studies of fiber-rich foods. However,
   from a scientific standpoint there is reason to believe that
   the use of fiberat higher intakes and higher concentration
   (i.e., a fiber supplement taken as a single dose prior to
   meals) may well differ from the typical intake of fiber from
   foods (i.e., smaller amounts of fiber consumed throughout the
   day). If an effect is seen when fiber is consumed at higher
   intakes and higher concentrations, it cannot be assumed that
   there is a linear dose-response effect that will translate
   into significant effects at more usual levels of intake.
   Conversely, if an effect is seen when fiber isconsumed with
   more typical intake from foods, it may not automatically
   translate into an effect from fiber supplements. Additionally,
   isolated fibers (as used in supplements) may vary in chemical
   and physical characteristics from native or less processed
   fibers. Since a mechanism of action for soluble fiber in
   reducing blood LDL-cholesterol levels has not been identified,
   it is possible that processing of soluble fibers may affect
   their ability to lower blood cholesterol levels.
   
   F. Comments Regarding Applicability to the General
   Population/Public Health Aspects
   
       42. One comment criticized FDA for including, in the
   dietary fiber-CVD proposal, the criterion that studies must be
   conducted in persons who generally represent the healthy U.S.
   population, i.e., adults with cholesterol readings below 300
   mg/dL. Thecomment stated that this criterion eliminated many
   valid studies that demonstrate significant scientific
   agreement for a claim about the relationship between soluble
   fiber and CHD; and secondly, that these studies remain useful
   to demonstrate soluble fiber's hypocholesterolemic effects at
   lower doses, for longer durations, in conjunction with
   reduced-fat diets.
       FDA agrees that populations at higher risk may provide a
   more sensitive group for identifying a nutrient/disease
   relationship. However, extrapolation of results from a high
   risk group to the general population must be done with caution
   and generallyrequires some confirmatory studies in the general
   population. FDA included, in its review, studies on persons
   considered to be at high risk, i.e., with levels of blood
   total cholesterol between 200 and 300 mg/dL. FDA, thus,
   included a large segment of the general population whose risk
   levels fall in this range, but excluded the much smaller
   segment of persons with blood total cholesterol levels above
   300 mg/dL, because these people often have multiple and
   serious health problems, making it difficult togeneralize
   results beyond the particular study populations. According to
   the National Center for Health Statistics (Ref. 79), only 5
   percent of the U.S. adult population of men and women have
   serum cholesterol levels 300 mg/dL or higher. Results of two
   studies (Refs. 9 and 59) showed that subjects with a mean
   serum cholesterol level >300 mg/dL had significantly lowered
   serum cholesterol at the end of the test period compared to
   baseline. Final serum cholesterol levels, however, was still
   greater than 300 mg/dL. Thus these individuals remained
   severely hypercholesterolemic and at high risk for CVD.
       It is also important to note that the hypocholesterolemic
   effect in these two studies (Refs. 9 and 59) and many of the
   other studies attenuated with time, and serum cholesterol
   levels increased toward baseline. Thus, the agency does not
   agree with thell compliance with the test regimen was
   accomplished.
       Of the other 33 studies reviewed in the proposed rule (56
   FR at 60596 through 60609), the mean blood total cholesterol
   was in the mid- to upper- 200's (mg/dL) because many of the
   subjects had individual serum cholesterol levels greater than
   300 mg/dL. These studies were included in the evaluation of
   the relationship between dietary fiber and CVD if they met the
   other evaluation criteria.
       43. Another comment criticized FDA for separately
   evaluating studies using ``typical'' or ``usual'' American
   diets (i.e., approximately 37 percent of calories from fat)
   and those using a reduced fat (Step 1) diet. The comment
   stated that the agency should have given equal weight to
   studies whether they involved a ``typical'' American diet or a
   reduced fat (Step 1) diet.
       FDA disagrees with this comment. As stated in the proposed
   rule (56 FR 60582 at 60587), ``Responses of blood cholesterol
   levels to dietary treatment are affected by many factors,
   including initial (baseline) blood cholesterol levels and
   dietary factors (i.e., the level of saturated fat and
   cholesterol in the diet).'' Because serum cholesterol is
   responsive to dietary intakes of saturated fat and
   cholesterol, FDA separated the studies on the basis of whether
   fiber effects were being evaluated as part of a typical
   American diet or as part of a reduced fat diet. Results of
   fiber studies become confounded when the test diet is not
   adequately controlled, or not assessed at all, and when
   subjects make their own changes to their diets by consuming
   less total and saturated fat. In such cases, the true effects
   of fiber, if any, cannot be adequately determined. Although
   the agency grouped studies based on type of diet, it did give
   them equal weight based on the evaluation criteria.
       44. One comment stated that FDA failed to consider studies
   reporting only modest reductions in serum cholesterol as
   strong evidence of soluble fiber's effectiveness. The comment
   refers to the public health significance of even a small
   reduction in serum cholesterol; i.e., a 1 percent reduction in
   serum cholesterol levels predicts approximately a 2 percent
   reduction in CHD (Ref. 41).
       FDA disagrees with this comment. FDA recognizes that
   changes in serum cholesterol in fiber-feeding studies are
   generally small. However, if, due to the sample size of the
   study, this change is not sufficient to demonstrate
   statistical significance beyond that which may occur by chance
   (generally accepted by epidemiologists and biostatisticians as
   p<.05), it cannot be concluded that a true effect has been
   observed.
       45. Another comment stated that there is no scientifically
   valid basis for excluding studies involving subjects whose
   blood cholesterol levels exceed 300 mg/dL.
       FDA disagrees with this comment. As stated above, 5
   percent of the U.S. population have serum cholesterol levels
   greater than 300 mg/dL (Ref. 79). It is important to remember
   that ``the magnitude of plasma lipid response is frequently
   related to the initial plasma lipid status of the experimental
   subjects. Persons with higher plasma lipids initially usually
   experience the greatest plasma lipid response to diet
   intervention'' (Ref. 33). What may cause an effect to look
   significant may be driven largely by the magnitude of the
   change, which is more likely to be observed in those
   individuals with very high serum cholesterol levels. To base
   conclusions on one group that will respond to a much greater
   extent than another is misleading as to those persons with
   normal to moderately elevated serum cholesterol levels, who
   comprise 95 percent of the adult population.
       46. One comment noted that population compliance was
   essential to the success of dietary intervention for improved
   public health and that adding fiber to the diet is more
   acceptable to consumers than removing fat.
       FDA agrees that compliance is an important factor, but,
   with respect to fiber and CVD, it must first be established
   that addition of fiber alone to the diet is an effective means
   of reducing serum cholesterol.
   
   G. Comments Regarding Issues of Study Design, Confounders,
   Fat, etc.
   
       47. One comment stated that FDA criticized studies that do
   not control for the effects of low dietary fat when examining
   increased dietary fiber consumption and cholesterol reduction.
       FDA disagrees with this comment. Although many studies
   provide compelling evidence of the effect of a combined
   low-fat, high fiber diet on lowering serum cholesterol, FDA
   examined the scientific evidence to determine whether a
   specific relationship existed between soluble dietary fiber
   and risk of heart disease, and whether sufficient scientific
   evidence was available to support a health claim for dietary
   fiber and heart disease.
       48. Another comment criticized FDA's evaluation of a study
   by Van Horn et al. (Ref. 67) for inappropriately discounting
   the correlation of fiber intake with serum cholesterol by
   pointing out that many other nutrient intakes besides fiber
   showed correlations with cholesterol levels. The comment
   stated that dietary fiber had correlations with CHD that were
   as statistically significant as those for fat and CHD, and
   this argues for a finding that significant scientific
   agreement supports a claim about the relationship between
   fiber and CHD.
       FDA disagrees with this comment. FDA again emphasizes that
   many other nutrient intakes in addition to dietary fiber show
   statistically significant correlations. In a cross-sectional
   survey analyzed in this manner, one of the first things that
   must be considered when many factors show significance is
   whether they are highly correlated with each other. If this is
   the case, displacement of fat in the diet by fiber is one
   explanation for the observed effect.
       49. Another comment argued that, although FDA has stated
   there may be other micronutrients or components in vegetables,
   cereal, fruits and berries other than soluble fiber that may
   have contributed to the cholesterol-lowering effects of some
   solublefiber-containing foods, the agency has failed to
   identify them.
       FDA agrees that it did not identify other components in
   foods that may have serum cholesterol lowering effects. Its
   point in this statement was that attribution of effects to
   soluble fiber per se, when foods contain a wide range of
   vitamins, minerals, and other substances, is misleading. The
   authors of the meta-analysis (Ref. 125) also noted that other
   components of oats may play a role in the cholesterol-lowering
   properties observed in human studies.
       50. Another comment criticized FDA's evaluation of a study
   by Van Horn et al. (Ref. 69).
       FDA stands by its evaluation of the Van Horn study. FDA
   criticized the study because there are significant changes in
   diet between the control and test group in total fat intake,
   saturated fat intake, and monounsaturated fat intake, and the
   authors did not demonstrate that these changes did not
   contribute to the cholesterol-lowering effect observed.
       51. Another comment suggested that weight loss is not a
   confounder, but rather a result of fiber's ability to effect
   weight reductions independently of changes in caloric or fat
   intake.
       FDA disagrees with this comment. Although this may be a
   possible mechanism of action, no studies reviewed have
   provided clear evidence of this.
       52. One comment stated that confounding variables and lack
   of a mechanism of action are not a valid basis for denying a
   link between ingestion of fiber and reduced risk of heart
   disease which experimental evidence shows in fact exists.
       FDA disagrees with the comment's description of the basis
   for its proposed action. If factors such as weight loss,
   changes in diet, lifestyle changes, and/or exercise, which are
   known to influence serum cholesterol, are not controlled in a
   clinical study, or if analytical methods for determining the
   soluble fiber content in food sources are not clearly
   established, the specificity for an effect of soluble fiber on
   serum cholesterol cannot be determined. In its proposal, FDA
   made a tentative decisionding dietary fiber and CVD for a
   number of reasons, including confounding factors identified in
   many of the clinical studies reviewed, but not because there
   is no recognized mechanism(s) of action for hypolipidemic
   effects of different soluble fibers. For a more thorough
   discussion of FDA's proposal to deny a health claim on dietary
   fiber and CVD in the proposed rule, see 56 FR 60582 at 60591
   through 60592. FDA believes that, for health claims to be
   educational and result in changed eating habits, the claims
   should be truthful and not misleading.
   
   H. Conclusions From Comments
   
       FDA agrees with many of the comments that a problem in
   determining the relationship between dietary fiber and heart
   disease is the fact that dietary fiber is a diverse group of
   chemical substances that may be associated with different
   physiological functions, and that the analytical methodology
   to identify the soluble fiber content of a food may not be
   predictive of the likely physiological effect.
   
   IV. Decision to Deny a Health Claim Relating Dietary Fiber to
   Reduced Risk of CHD
   
       Overall, the currently available scientific evidence is
   not sufficiently conclusive or specific for soluble fiber per
   se to justify use of a health claim relating the intake of
   dietary soluble fiber to a reduced risk of heart disease. A
   major limitation in designing and evaluating research studies
   has been the need for better defined measures of dietary
   soluble fiber and standardized descriptions for source, type,
   and amount of dietary soluble fiber. Commonly used analytical
   methodologies do not detect many of the characteristics that
   may vary among fibers and that may be related to biological
   function (e.g., particle size, chemical composition,
   water-holding capacity). Other components associated with
   soluble fiber in foods (e.g., tocotrienols) may also have some
   cholesterol-lowering capabilities. The inability to detect
   many of the differences among fibers, fiber components, and
   other substances in foods which contain soluble fiber, and the
   general lack of conclusions as to the mechanism(s) of action
   of soluble fibers raise questions about the ability of
   commonly used analytical measures of dietary fiber to
   adequately predict biological actions of specific fibers. For
   these reasons, FDA is not authorizing use of a health claim
   relating dietaryfiber to a decreased risk of CHD.
       FDA's decision is consistent with recent conclusions
   reached about the state of the scientific evidence by the
   National Heart, Lung, and Blood Institute of the National
   Institutes of Health (Ref. 155) and recommendations in the
   Institute of Medicine's report ``Nutrition Labeling: Issues
   and Directions for the 1990s'' (Ref. 81). This report notes
   that there has been great interest in the specific effects of
   dietary fiber on several chronic diseases. According to the
   report, the strongest argument for an increased consumption of
   dietary fiber is the important contribution it makes to normal
   bowel function. Clear scientific associations of fiber intake
   with the incidence of cancer, heart disease, and diabetes
   mellitus have not been made. The report indicates that one
   reason for this may be the difficulty in designing appropriate
   experiments to test specifically for the effect of dietary
   fiber. Foods high in dietary fiber are also generally low in
   calories and total and saturated fatty acids and devoid of
   cholesterol; thus, determination of a specific fiber effect in
   a feeding study is difficult. Moreover, according to the
   report, foods have a variety of fiber components and each may
   have different actions. Chemically and physiologically,
   cellulose,lignin, hemicellulose, pectin, and alginate (all
   relatively purified fiber types) behave differently. Wheat
   bran, oat bran, and rice bran (all heterogeneous mixtures of
   fibers) are not similar in composition. It is also very
   difficult to analyze dietary fnd labeling of foods of health
   claims relating to an association between the ingestion of
   dietary fiber (particularly soluble fiber) and a reduction in
   the risk of heart disease. In reaching this decision, the
   agency considered all comments received in response to its
   proposed rule (56 FR 60582), and reviewed the scientific
   literature that became publicly available after the proposal's
   publication and data submitted with comments.
   
   V. Decision to Allow a Health Claim on Foods Relating Diets
   Low in Saturated Fat and Cholesterol and High in Fruits,
   Vegetables, and Grain Products, Foods That Contain Fiber,
   Particularly Soluble Fiber, to a Reduced Risk of CHD
   
       FDA has reviewed the numerous authoritative documents,
   including Federal government reports, as well as recent
   research on dietary fiber and CHD risk. In addition, the
   agency considered all comments received in response to its
   proposed rule. The agency has concluded that the publicly
   available scientific evidence supports an association between
   diets low in saturated fat and cholesterol and high in fruits,
   vegetables, and grain products, foods that are low in
   saturated fat and cholesterol and are good sources of dietary
   fiber, to reduced risk of heart disease.
       FDA agrees with the comments that show that dietary
   patterns that are low in saturated fat and cholesterol and
   high in fruits, vegetables (including legumes), and grain
   products are associated with a decreased risk of CHD. Although
   the specific roles of the numerous potentially protective
   substances in such plant foods are not yet understood,
   populations with diets rich in these foods experience many
   health advantages, including lower rates of heart disease.
   Currently, there is not scientific agreement as to whether the
   observed protective effects against heart disease are due to a
   combination of nutrient components of the foods, including
   soluble fiber, to other components of soluble fiber-rich diets
   (for example, potassium and magnesium), to displacement of
   saturated fat and cholesterol from the diet, or to
   non-nutritive substances in these foods.
       Thus, the conclusion that diets low in saturated fat and
   cholesterol and high in fruits, vegetables, and grain
   products, foods low in saturated fat and cholesterol and
   containing soluble fiber, are associated with a reduced risk
   of heart disease is consistent with the available scientific
   evidence. The fact that these foods contain dietary fiber,
   particularly soluble fiber, can serve, therefore, as a useful
   marker for identifying those fruits, vegetables, and grain
   products which, when added to diets low in saturated fat and
   cholesterol, may help in reducing blood LDL-cholesterol
   levels. As discussed in the final rule on general requirements
   for health claims, published elsewhere in this issue of the
   Federal Register, statements about good nutrition that do not
   expressly or by implication refer to a substance are
   considered dietary guidance and not health claims. In this
   rule, FDA is authorizing the inclusion of a reference to
   dietary fiber (a substance) in a statement about the value of
   fruit, vegetables, and grain products in reducing the risk of
   heart disease. Thus, the health claim permitted under this
   regulation to be used on labels and labeling of certain foods
   associates diets low in saturated fat and cholesterol and high
   in vegetables, fruit, and grain products, that contain soluble
   fiber with a reduced risk of heart disease.
   
   VI. Rationale and Description of the Final Regulation
   
   A. Relationship and Significance Statements
   
       In new 101.77(a), the summary of the relationship between
   diets low in saturated fat and cholesterol and high in fruits,
   vegetables, and grain products that contain soluble fiber and
   reduced heart disease risk is consistent with the conclusions
   reached in the review of the scientific evidence. Although the
   specific roles of dietary soluble fiber, or of specific
   soluble fibers and fiber components, and the multiple
   nutrients and other substances contained in these foods, are
   not yet fully understo many studies have shown that diets high
   in soluble-fiber-containing foods are associated with lower
   blood LDL-cholesterol levels and with reduced risk of heart
   disease. These diets are generally low in saturated fat,
   cholesterol and total fat, nutrients known to have a
   detrimental effect on blood LDL-cholesterol levels, and
   therefore, on risk of heart disease. Dietary soluble fiber can
   be used as a marker to identify the types of foods which
   correlate with reduced heart disease risk, and whose addition
   to diets low in saturated fat and cholesterol is considered to
   be useful in lowering blood LDL-cholesterol (Ref. 66). The
   relationship statement in 101.77(a) also includes other
   information about heart disease, such as risk factors, as in
   other authorized health claims.
       New 101.77(b), on the significance of the relationship
   between consumption of diets low in saturated fat and
   cholesterol and high in fruits, vegetables, and grain products
   that contain soluble fiber and reduced risk of heart disease,
   includes the information that U.S. diets tend to be high in
   saturated fat and cholesterol and low in fiber-containing
   fruits, vegetables, and grain products. A discussion of
   current dietary guidelines on recommended servings of fruits,
   vegetables, and grain products is also provided.
   
   B. Nature of the Claim
   
       In new 101.77(c)(2)(i), FDA is authorizing a health claim
   relating diets low in saturated fat and cholesterol and high
   in fruits, vegetables, and grain products that contain soluble
   fiber to reduced risk of heart disease. In new
   101.77(c)(2)(i)(A), the agency is requiring, consistent with
   other authorized claims, that the relationship be qualified
   with the terms ``may'' or ``might.'' These terms are used to
   indicate that not all persons can necessarily expect to
   benefit from these dietary changes.
       In new 101.77(c)(2)(i)(B), the agency, consistent with
   other authorized claims, is requiring that the claim use the
   specific terms ``heart disease'' or ``coronary heart disease''
   to define the type of disease dealt with by this claim. These
   disease terms reflect terms commonly used in dietary guidance
   materials, and are also reflective of the scientific evidence
   which links these dietary factors to heart disease risk via
   the intermediate mechanism of reducing blood LDL-cholesterol
   levels, rather than to the broader category of cardiovascular
   disease.
       In new 101.76(c)(2)(i)(C), the agency is requiring that
   the claim discuss only those fruits, vegetables, and grain
   products that contain dietary fiber, rather than all fruits,
   vegetables, and grain products. Diets low in saturated fat and
   cholesterol and high in fruits, vegetables, and grain products
   that contain fiber, particularly soluble fiber, are correlated
   with reduced heart disease risk. Thus, a claim discussing
   those fruits, vegetables, and grain products that contain the
   marker nutrient, but does not attribute a protective effect to
   soluble fiber, is consistent with current scientific
   knowledge.
       New 101.77(c)(2)(i)(D) specifies the terms that can be
   used to describe the fiber component of the fruits,
   vegetables, and grain products that bear a health claim.
   Consistent with the state of the scientific evidence, this
   paragraph permits a choice among a number of terms, but does
   not allow terms for specific types of fiber to be used, e.g.,
   those connoting the origin of the fiber. The term ``soluble
   fiber'' may be used in combination with a more general term
   for fiber. This permits referenceto soluble fiber, which is a
   useful marker nutrient for foods associated with reduced risk
   of heart disease. However, the present scientific evidence
   does not permit a determination of whether it is the soluble
   fiber or other components in these foods or displacement of
   fat that provides the protective effect. Given these
   uncertainties about the specific role of soluble fiber, it
   would be misleading to place undue emphasis on soluble fiber
   standing alone.
       New 101.77(c)(2)(i)(F) requires that health claims
   specify that development of heart disease depends on many
   factors. This requirement is intended to prevent consumers
   from being misled that fruits, vegetables, and grain product
   intake is the only factor connected with heart disease risk.
       In new 101.77(c)(2)(i)(E), FDA is prohibiting, consistent
   with other authorized health claims, more specific use of
   dietary terms than is warranted by the current state of the
   scientific evidence. New 101.77(c)(2)(i)(G) is also
   consistent with other authorized health claims, and prohibits
   any attribution of degree of risk for heart disease and
   fiber-containing fruits, vegetables, and grains. These
   requirements also standardize use of terms, thus minimizing
   consumer confusion as they comparefood labels across products,
   or as they compare a health claim to the nutrition information
   panel.
   
   C. Nature of the Food
   
       New 101.77(c)(2)(ii)(A) requires that the food bearing
   the health claim be or contain a fruit, vegetable, or grain
   product. Because the claim relates to diets high in these
   foods, it would not make sense for it to appear on the
   labeling of another type of food. A health claim that appears
   on a food that meets all the requirements in
   101.77(c)(2)(ii), but contains only a trivial amount of
   fruit, vegetable, or grain product, could be considered
   misleading and might misbrand the food under section 403(a) of
   the act.
       FDA, consistent with the requirements for the health claim
   on dietary saturated fat and cholesterol and heart disease
   (published elsewhere in this issue of the Federal Register),
   is requiring in new 101.77(c)(2)(ii)(B) that foods bearing
   the health claim meet requirements for ``low saturated fat,''
   ``low cholesterol,'' and ``low total'' fat, or alternatively,
   belong to a class of products that is ``low in saturated
   fat,'' ``low in cholesterol,'' and ``low in total fat.'' Low
   saturated fatand cholesterol diets are associated with reduced
   heart disease risk. Low or negligible total fat is also one of
   the characterizing features of diets rich in fiber-containing
   fruits, vegetables, and grain products. Because the effects of
   saturated fat and cholesterol are not readily separated from
   the effects of other nutritive components of fruits,
   vegetables, and grain products, and because the scientific
   evidence linking diets low in saturated fat and cholesterol to
   reduced risk of heart disease is strong, saturated fat and
   cholesterol are specified as qualifying nutrients. Total fat
   is also specified as a qualifying nutrient because a low
   content of total fat is characteristic of dietary patterns
   which relate to lower heart disease risk, and because it
   facilitates the ability of consumers to achieve diets low in
   saturated fat and cholesterol. (See final rule on ``Dietary
   Saturated Fat and Cholesterol and Coronary Heart Disease,''
   published elsewhere in this issue of the Federal Register).
       In new 101.77(c)(2)(ii)(C), FDA is requiring that fruits,
   vegetables, and grain products bearing the authorized health
   claim contain at least 0.6 g of dietary soluble fiber per
   reference amount commonly consumed. Because soluble fiber is a
   qualifying nutrient, FDA, in new 101.77(c)(2)(ii)(D), is
   requiring declaration of soluble fiber content consistent with
   requirements in 101.9(c)(6)(i)(A). The qualifying value of
   0.6 g of soluble fiber is based on several considerations.
   First, an expert panel convened by LSRO (Ref. 39) recommended
   total dietary fiber intakes of 20 to 30 g daily for adults
   (see final rule on daily reference values published elsewhere
   in this issue of the Federal Register). It further recommended
   that approximately 25 percent (or about 6 g) of this be
   soluble fiber. This level of soluble fiber represents the same
   ratio of soluble to insoluble fiber normally found in foods
   and for which there is a long history of use, and therefore
   was considered by the panel to bed to double their intakes to
   meet the current dietary guidelines. A total daily intake of 6
   g of soluble fiber from fruits, vegetables, and grains is
   consistent with current dietary guidelines for the general
   population.
       The qualifying criterion of 0.6 g per reference amount
   customarily consumed is also consistent with the definition of
   a ``good source'' of a nutrient (i.e., 10 percent of the daily
   reference value (DRV)) in the final rule on general
   requirements fornutrient content claims published elsewhere in
   this issue of the Federal Register. Although there is no DRV
   for soluble fiber, the requirement that a nutrient be present
   at 10 percent of a reference standard has been set as a
   qualifying level in other regulations authorizing health
   claims. (See the final rules on antioxidant vitamins and
   cancer and on fiber cancer published elsewhere in this issue
   of the Federal Register.) The 10 percent level is deemed
   useful and appropriate, because very few foods could naturally
   meet the requirement for a ``high'' source of soluble fiber.
   The current dietary guidance recommendations of five or more
   servings of fruits and vegetables and six or more servings of
   grain products daily, if followed, would likely result in
   intakes of soluble fiber close to or exceeding the recommended
   daily intake of 6 g. Thus, use of a qualifying criterion
   consistent with that used to define a ``good'' source for
   nutrients which have DRV's provides for an amount that allows
   a number of fruits, vegetables, and grain products to qualify,
   and is consistent with current dietary guidelines for general
   dietary patterns. Without this alternate level, very few
   fruits, vegetables, and grain products would qualify for the
   health claim, which would be contrary to the available
   scientific evidence and to the purpose of health claims.
       Section 101.77(c)(2)(ii)(C) also requires that foods
   qualify as a good source of soluble fiber based on their
   natural level of soluble fiber. This means that foods which
   require fortification with soluble fiber, in order to meet the
   qualifying criteria for the health claim, cannot bear the
   claim. This requirement is consistent with the scientific
   basis for the claim, that is, that intakes of fruits,
   vegetables, and grains in their native form correlate with
   reduced heart disease risk. Because thereare not sufficient
   data that specifically identify dietary soluble fiber, or
   particular components of soluble fiber, as the cause of a
   reduction in heart disease risk, and because this nutrient is
   being used as a marker for the substance or substances in
   fruits, vegetables, and grain products that provide the
   observed protective effect, it is the native composition of
   the foods that identifies their usefulness.
   
   D. Optional Information
   
       Under new 101.77(d), similarly to other authorized health
   claims, health claims may identify additional risk factors for
   heart disease. The regulation specifies the factors that may
   be listed; all are risk factors about which there is general
   scientific agreement. This additional information can provide
   a context that is useful for an understanding of the
   relationship of the diet to the disease, but manufacturers are
   cautioned that it should not be presented in a way that is
   misleading to the consumer. A health claim may also indicate
   that reductions in saturated fat and cholesterol intake and
   consumption of fruits, vegetables, and grain products are part
   of a total dietary pattern that is consistent with the latest
   ``Nutrition and Your Health: Dietary Guidelines for
   Americans,'' published jointly by the U.S. Department of
   Agriculture and the U.S. Department of Health and Human
   Services. Consistent with other health claim regulations, the
   claim may also include information on the prevalence of heart
   disease in the United States. In order to ensure that this
   information is valid, the agency is requiring that it come
   from one of three specified authoritative sources. Finally,
   because consumers frequently know their cholesterol levels or
   can detegh readily available facilities in shopping malls and
   health clinics, the agency, similarly to other authorized
   health claims, is requiring that when information in the claim
   allows consumers to ``self-diagnose'' their own risk level,
   that additionally,the claim indicate the need for medical
   guidance if a consumer falls within a risk category.
       Similarly to the requirements in 101.73 on ``Dietary
   Saturated Fat and Cholesterol and Coronary Heart Disease,''
   the claim may indicate that the relationship between diets low
   in saturated fat and cholesterol and high in fruits,
   vegetables, and grains that contain fiber, particularly
   soluble fiber, is through the intermediate link of ``blood
   cholesterol'' or ``blood total- and LDL-cholesterol.'' Such
   information is useful to consumers, but could add
   unnecessarily to the length and complexity ofthe required
   health claim. For these reasons, this provision is optional
   rather than mandatory.
   
   E. Model Health Claims
   
       In new 101.77(e), FDA is providing model health claims to
   illustrate the requirements of new 101.77. FDA is not
   prescribing specific language for claims, but certain elements
   are required, and these models include the required elements.
   
   VII. Environmental Impact
   
       The agency has determined under 21 CFR 25.24(a)(11) 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.
   
   VIII. Economic Impact
   
       In its food labeling proposals of November 27, 1991 (56 FR
   60366 et seq.), FDA stated that the food labeling reform
   initiative, taken as a whole, would have associated costs in
   excess of the $100 million threshold that defines a major
   rule. Thus, inaccordance with Executive Order 12291 and the
   Regulatory Flexibility Act (Pub. L. 96 - 354), FDA developed
   one comprehensive regulatory impact analysis (RIA) that
   presented the costs and benefits of all of the food labeling
   provisions taken together. That RIA was published in the food
   labeling proposals of November 27, 1991 (56 FR 60366 et seq.),
   FDA stated that the food labeling reform initiative, taken as
   a whole, would have associated costs in excess of the $100
   million threshold that defines a major rule. Thus, in
   accordance with Executive Order 12291 and the Regulatory
   Flexibility Act (Pub. L. 96 - 354), FDA developed one
   comprehensive regulatory impact analysis (RIA) that presented
   the costs and benefits of all of the food labeling provisions
   taken together. That RIA was published in the Federal Register
   of November 27, 1991 (56 FR 60856), and along with the food
   labeling proposals, the agency requested comments on the RIA.
       FDA has evaluated more than 300 comments that it received
   in response to the November 1991 RIA. FDA's discussion of
   these comments is contained in the agency's final RIA
   published elsewhere in this issue of the Federal Register. In
   addition, FDA will prepare a final regulatory flexibility
   analysis (RFA) subsequent to the publication of the food
   labeling final rules. The final RFA will be placed on file
   with the Dockets Management Branch (HFA - 305), Food and Drug
   Administration, rm. 1 - 23, 12420 Parklawn Dr., Rockville, MD
   20857, and a notice will be published in the Federal Register
   announcing its availability.
       In the final RIA, FDA has concluded, based on its review
   of available data and comments, that the overall food labeling
   reform initiative constitutes a major rule as defined by
   Executive Order 12291. Further, the agency has concluded that
   although the costs of complying with the new food labeling
   requirements are substantial, such costs are outweighed by the
   public health benefits that will be realized through the use
   of improved nutrition information provided by food labeling.
   
   IX. References
   
       The following references have been placed 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. Abraham, Z. D., and T. Mehta, ``Three-week Psyllium
   Supplementation: Effect on Plasma Cholesterol Concentrations,
   Fecal Steroid Excretion, and Carbohydrate Absorption in Men,''
   American Journal of Clinical Nutrition, 47:67 - 74, 1988.
       2. Anderson, J. W., D. A. Deakins, T. L. Floore, B. M.
   Smith, and S. E. Whitis,``Dietary Fiber and Coronary Heart
   Disease,'' in: Clydesdale, F. M., ed., Critical Reviews in
   Food Science and Nutrition, Vol. 29, New York, pp. 95 - 137,
   1990.
       3. Anderson, J. W., N. H. Gilinsky, D. A. Deakins, S. F.
   Smith, D. S. O'Neal, D. W. Dillon, P. R. Oeltgen, ``Lipid
   Responses of Hypercholesterolemic Men to Oat-bran and Wheat
   bran Intake,'' American Journal of Clinical Nutrition, 54:678
   - 83, 1991.
       4. Anderson, J. W., N. J. Gustafson, D. B. Spencer, J.
   Tietyen, C. A. Bryant, ``Serum Lipid Response of
   Hypercholesterolemic Men to Single and Divided Doses of Canned
   Beans,'' American Journal of Clinical Nutrition, 51:1013 - 9,
   1990.
       5. Anderson, J. W., S. Riddell-Lawrence, T. L. Floore, D.
   W. Dillon, P. R. Oeltgen, ``Effects of Refined Bakery Products
   on Serum Cholesterol Levels in Hypercholesterolemic Men,''
   unpublished, 1991.
       6. Anderson, J. W., D. B. Spencer, C. C. Hamilton, S. F.
   Smith, J. Tietyen, C. A. Bryant, P. Oeltgen,``Oat-bran Cereal
   Lowers Serum Total and LDL Cholesterol in Hypercholesterolemic
   Men,'' American Journal of Clinical Nutrition, 52:495 - 499,
   1990.
       7. Anderson, J. W., L. Story, B. Sieling, W. J. L. Chen,
   M. S. Petro, J. Story, ``Hypocholesterolemic Effects of
   Oat-bran or Bean Intake for Hypercholesterolemic Men,''
   American Journal of Clinical Nutrition, 40:1146 - 55, 1984.
       8. Anderson, J. W., N. Zettwoch, T. Feldman, J.
   Tietyen-Clark, P. Oeltgen, C. W. Bishop,
   ``Cholesterol-lowering Effects of Psyllium Hydrophilic
   Mucilloid for Hypercholesterolemic Men,''Archives of Internal
   Medicine, 148:292 - 296, 1988.
       9. Aro, A., M. Uusitupa, E. Voutilainen, T. Korhonen,
   ``Effects of Guar Gum in Male Subjects With
   Hypercholes-terolemia,'' American Journal of Clinical
   Nutrition, 39:911 - 916, 1984.
       10. Beling, S., L. Detrick, W. Castelli, ``Serum
   Cholesterol Response to a Processed Oat Bran Cereal Among
   Hypercholes-terolemics on a Fat-modified Diet,'' unpublished
   clinical trial submitted by the Quaker Oats Co., 1991.
       11. Bell, L. P., K. J. Hectorn, Reynolds, H., Balm, T. K.,
   Hunninghake, D. B. ``Cholesterol-lowering Effects of Psyllium
   Hydrophilic Mucilloid--Adjunct Therapy to a Prudent Diet for
   Patients With Mild to Moderate Hypercholesterolemia,'' Journal
   ofthe American Medical Association, 261:3419 - 3423, 1989.
       12. Bell, L. P., K. J. Hectorn, H. Reynolds, D. B.
   Hunninghake, ``Cholesterol-Lowering Effects of Soluble Fiber
   Cereals as Part of a Prudent Diet for Patients With Mild to
   Moderate Hypercholesterolemia,'' American Journal of Clinical
   Nutrition, 52:1020 - 6, 1990.
       13. Burr, M. L., J. F. Gilbert, R. M. Holliday, P. C.
   Elwood, A. M. Fehily, S. Rogers, P. M. Sweetnam, N. M.
   Deadman, ``Effects of Changes in Fat, Fish, and Fiber Intakes
   on Death and Myocardial Reinfarction: Diet and Reinfarction
   Trial (DART),'' Lancet, pp. 757 - 761, September 30, 1989.
       14. Consensus Development Panel, ``Lowering Blood
   Cholesterol to Prevent Heart Disease,'' Journal of the
   American Medical Association, 253:2080 - 86, 1985.
       15. Davidson, M. H., L. D. Dugan, J. H. Burns, J. Bova, K.
   Story, K. B. Drennan, ``The Hypocholesterolemic Effects of
   Beta-glycan in Oatmeal and Oat Bran--A dose-controlled
   Study,'' Journal of the American Medical Association,
   265(14):1833 - 39, 1991.
       16. Demark-Wahnefried, W., J. Bowering, P. S. Cohen,
   ``Reduced Serum Cholesterol With Dietary Change Using
   Fat-modified and Oat Bran Supplemented Diets,'' Journal of the
   American Dietetic Association, 90:223 - 9, 1990.
       17. Furda, I., ``Interaction of Dietary Fiber with
   Lipids--
   Mechanistic Theories and Their Limitations,'' in: Furda, I.,
   C. J. Brine, eds., New Developments in Dietary Fiber, New
   York: Plenum Press. p. 67 - 80, 1990.
       18. Glassman, M., Spark, A., Berezin, S., Schwarz, S.,
   ``Treatment of Type Iia Hyperlipidemia in Childhood by a
   Simplified American Heart Association Diet and Fiber
   Supplementation,'' American Journal of Diseases of Children,
   144:973 - 76, 1990.
       19. Gold, K. V., D. M. Davidson, ``Oat Bran as a
   Cholesterol-reducing Dietary Adjunct in a Young, Healthy
   Population,'' Western Journal of Medicine, 148:299 - 302,
   1988.
       20. Grande, F., J. T. Anderson, A. Keys, ``Effect of
   Carbohydrates of Leguminous Seeds, Wheat and Potatoes on Serum
   Cholesterol Concentration in Man,'' Journal of Nutrition,
   86:313 - 317, 1965.
       21. Grande, F., J. T. Anderson, A. Keys, ``Sucrose and
   Various Carbohydrate-Containing Foods and Serum Lipids in
   Man,'' American Journal of Clinical Nutrition, 27:1043 - 1051,
   1974.
       22. Jenkins, D. J. A., A. R. Leeds, C. Newton, J. H.
   Cummings, ``Effect of Pectin, Guar Gum, and Wheat Fiber on
   Serum Cholesterol,'' Lancet, i:1116 - 1117, 1975.
       23. Jenkins, D. J. A., D. Reynolds, A. R. Leeds, A. L.
   Waller, J. H. Cummings, ``Hypocholesterolemic Action of
   Dietary Fiber Unrelated to Fecal Bulking Effect,'' American
   Journal of Clinical Nutrition, 32:2430 - 2435, 1979.
       24. Judd, P. A., A. S. Truswell, ``Dietary Fibre and Blood
   Lipids in Man,'' in: ``Dietary Fiber Perspectives: Reviews and
   Bibliography,'' A. R. Leeds, A. Avenell, eds, London: John
   Libbey and Co., pp. 23 - 39, 1985.
       25. Kahn, R. F., K. W. Davidson, J. Garner, R. S. McCord,
   ``Oat Bran Supplementation for Elevated Serum Cholesterol,''
   Family Practice Research Journal, 10:37 - 46, 1990.
       26. Keenan, J. M., J. B. Wenz, S. Myers, C. Ripsin, Z.
   Huang, ``Randomized Controlled Cross-over Trial of Oat Bran in
   Hypercholes-terolemic Subjects,'' in Press, 1991.
       27. Kesaniemi, Y. A., S. Tarpila, T. A. Miettinen, ``Low
   versus High Dietary Fiber and Serum, Biliary, and Fecal Lipids
   in Middle aged Men,'' American Journal of Clinical Nutrition,
   51:1007 - 12, 1990.
       28. Kestin, M., R. Moss, P. M. Clifton, P. J. Nestel,
   ``Comparative Effects of Three Cereal Brans on Plasma Lipids,
   Blood Pressure, and Glucose Metabolism in Mildly
   Hypercholesterolemic Men,'' American Journal of Clinical
   Nutrition, 52:661 - 6, 1990.
       29. Keys, A., J. T. Anderson, F. Grande, ``Diet-type (Fats
   Constant) and Blood Lipids in Man,'' Journal of Nutrition,
   70:257 - 266, 1960.
       30. ``Coronary Heart Disease in Seven Countries,'' Keys,
   A., Ed., Circulation, 41 (Suppl. 1).
       31. Khaw, K., E. Barret-Connor, ``Dietary Fiber and
   Reduced Ischemic Heart Disease Mortality Rates in Men and
   Women: A 12-year Prospective Study,'' American Journal of
   Epidemiology, 126:1093 - 1102, 1987.
       32. Kirby, R. W., J. W. Anderson, B. Sieling, E. D. Rees,
   W. J. L. Chen, R. E. Miller, R. M. Kay, ``Oat-bran Intake
   Selectively Lowers Serum Low-density Lipoprotein Cholesterol
   Concentrations of Hypercholesterolemic Men,'' American Journal
   of Clinical Nutrition, 34:824 - 829, 1981.
       33. Kris-Etherton, P. M., D. Krummel, M. E. Russell, D.
   Dreon, S. Mackey, J. Borchers, P. D. Wood, ``National
   Cholesterol Education Program--The Effect of Diet on Plasma
   Lipids, Lipoproteins, and Coronary Heart Disease,'' Journal of
   the American Dietetic Association, 88:1373 - 1400, 1988.
       34. Kritchevsky, D., ``Lipid Metabolism and Coronary Heart
   Disease,'' in: ``Dietary Fibre, Fibre-depleted Foods and
   Disease,'' Trowell, H., D. Burkitt, K. Heaton, Eds., New York:
   Academic Press. pp. 305 - 313, 1985.
       35. Lampe, J. W., J. L. Slavin, K. S. Baglien, W. O.
   Thompson, W. C. Duane, J. H. Savoral, ``Serum Lipid and Fecal
   Bile Acid Changes with Cereal, Vegetable, and Sugar-beet Fiber
   Feeding,'' American Journal of Clinical Nutrition, 53:1235 -
   1241,
       36. Lanza, E., D. Y. Jones, G. Block, L. Kessler,
   ``Dietary Fiber Intake in the U. S. Population,'' American
   Journal of Clinical Nutrition, 46:790 - 797, 1987.
       37. Levin, E. G., V. T. Miller, R. A. Muesing, D. B. Stoy,
   T. K. Balm, J. C. LaRosa, ``Comparison of Psyllium Hydrophilic
   Mucilloid and Cellulose as Adjuncts to a Prudent Diet in the
   Treatment of Mild to Moderate Hypercholesterolemia,'' Archives
   of Internal Medicine, 150:1822 - 27.
       38. ``Nutrition, Lipids, and Coronary Heart Disease: A
   Global View,'' Levy, R., B. Rifkind, B. Dennis, N. Ernst,
   Eds., Raven, New York, 1979
       39. LSRO, FASEB, ``Physiological Effects and Health
   Consequences of Dietary Fiber,'' Bethesda, MD, 1987.
       40. LSRO, FASEB, ``Dietary Fiber and Cardiovascular
   Disease,'' Bethesda, MD, 1991.
       41. Lipids Research Clinics Program, ``The Lipid Research
   Clinics Coronary Primary Prevention Trial Results II,'' in The
   Relationship of Reduction in Incidences of Coronary Heart
   Disease to Cholesterol Lowering, Journal of the Aerican
   Medical Association, 251:365 - 74, 1984.
       42. Little, P., G. Girling, A. Craven, A. Trafford, ``The
   Effect of a Combination Low Sodium, Low Fat, High Fibre Diet
   on Serum Lipids in Treated Hypertensive Patients,'' European
   Journal of Clinical Nutrition, 44:293 - 300, 1990.
       43. Lo, G. S., T. G. Cole, ``Soy Cotyledon Fiber Products
   Reduce Plasma Lipids'' Atherosclerosis, 82:59 - 67, 1990.
       44. McIntosh, G. H., J. Whyte, R. McArthur, P. J. Nestel,
   ``Barley and Wheat Foods: Influence on Plasma Cholesterol
   Concentrations in Hypercholesterolemic Men,'' American Journal
   of Clinical Nutrition, 53:1205 - 1209, 1991.
       45. McIvor, M. E., C. C. Cummings, A. I. Mendeloff,
   ``Long-term Ingestion of Guar Gum is Not Toxic in Patients
   with Noninsulin-Dependent Diabetes Mellitus,'' American
   Journal of Clinical Nutrition, 41:891 - 894, 1985.
       46. Minister of National Health and Welfare Canada, Report
   of the Expert Advisory Committee on Dietary Fibre, Canadian
   Government Publishing Centre, Ottawa, Canada, 1985.
       47. Minister of National Health and Welfare Canada,
   ``Nutrition Recommendations: The Report of the Scientific
   Review Committee 1990,'' Canadian Government Publishing
   Centre, Ottawa, Canada, 1990.
       48. National Research Council/National Academy of Sciences
   (NAS), ``Diet and Health: Implications for Reducing Chronic
   Disease Risk,'' National Academy Press: Washington, DC, 1989.
       49. National Research Council (NRC). ``Recommended Dietary
   Allowances,'' 10th Edition, National Academy Press:
   Washington, DC, 1989.
       50. Newman, R. K., S. E. Lewis, C. W. Newman, R. J. Boik,
   R. T. Ramage, ``Hypocholesterolemic Effect of Barley Foods on
   Healthy Men,'' Nutritional Reports International, 39:749 -
   760, 1989.
       51. Nishina, P. M., R. A. Freedland, ``Effects of
   Propionate on Lipid Biosynthesis in Isolated Rat Hepatocytes''
   Journal of Nutrition, 120:668 - 673, 1990.
       52. Ross, R., ``Atherosclerosis'' in: Cecil Textbook of
   Medicine, J. B. Wyngaarden, L. H. Smith, eds., W. B. Saunders
   Co., Philadelphia, PA, pp. 319 - 320, 1988.
       53. Shutler, S. M., G. M. Bircher, J. A. Tredger, L. M.
   Morgan, A. F. Walker, A. G. Low, ``The Effect of Daily Baked
   Bean (Phaseolus Vulgaris) Consumption on the Plasma Lipid
   Levels of Young, Normo-cholesterolaemic Men,'' British Journal
   ofNutrition, 61:257 - 65, 1989.
       54. Stamler, J., ``Nutrition-related Risk Factors For The
   Atherosclerotic Diseases--Present Status,'' Progress in
   Biochemical Pharmacology, 19:245 - 308, 1983.
       55. Stewart, R. B., W. E. Hale, M. T. Moore, F. E. May, R.
   G. Marks, ``Effect of Psyllium Hydrophilic Mucilloid on Serum
   Cholesterol in the Elderly'' Digestive Disease and Sciences,
   36:329 - 334, 1991.
       56. Superko, H. R., W. L. Haskell, L. Sawrey-Kubicek, J.
   W. Farquhar,``Effects of Solid and Liquid Guar Gum on Plasma
   Cholesterol and Triglyceride Concentrations in Moderate
   Hypercholesterolemia,'' American Journal of Cardiology, 62:51
   - 55, 198
       57. Swain, J. F., I. L. Rouse, C. B. Curley, F. M. Sacks,
   ``Comparison of the Effects of Oat Bran and Low fiber Wheat on
   Serum Lipoprotein Levels and Blood Pressure'' New England
   Journal of Medicine, 322:147 - 52, 1990.
       58. Taneja, A., C. M. Bhat, A. Arora, A. P. Kaur, 1989.
   ``Effect of Incorporation of Isabgol Husk in a Low Fibre Diet
   on Faecal Excretion and Serum Levels of Lipids in Adolescent
   Girls,'' European Journal of Clinical Nutrition, 43:197 - 202,
   1989.
       59. Tuomilehto, J., M. Silvasti, A. Aro, P. Karttunen, C.
   Gref, C. Ehnholm, M. Uusitupa, ``Long Term Treatment of Severe
   Hypercholesterolaemia With Guar Gum,'' Atherosclerosis, 72:157
   - 162, 1988.
       60. Tuomilehto, J., E. Voutilainen, J. Huttunen, S. Vinni,
   K. Homan,``Effect of Guar Gum on Body Weight and Serum Lipids
   in Hypercholesterolemic Females,'' Acta Medica Scandinavian,
   208:45 - 48, 1980.
       61. I. H. Ullrich, ``Evaluation of a High-fiber Diet in
   Hyperlipidemia: A Review,'' Journal of the American College of
   Nutritton, 6(1):19 - 25, 1987.
       62. USDA and DHHS, ``Nutrition and Your Health: Dietary
   Guideline for Americans''' 3rd Ed., Home and Garden Bulletin
   No. 232, U.S. Government Printing Office, Washington, DC.
       63. DHHS, Public Health Service (PHS), ``The Surgeon
   General's Report on Nutrition and Health,'' U.S. Government
   Printing Office, Washington, DC, 1988.
       64. DHHS, PHS, ``Healthy People 2000: National Health
   Promotion and Disease Prevention Objectives,'' Washington, DC,
   1990.
       65. DHHS, PHS and the National Institutes of Health (NIH),
   ``National Cholesterol Education Program: Report of the Expert
   Panel on Detection, Evaluation, and Treatment of High Blood
   Cholesterol in Adults,'' NIH, Bethesda, MD, 1989.
       66. DHHS, PHS and the National Institutes of Health,
   ``National Cholesterol Education Program: Population Panel
   Report,'' Bethesda, MD, 1989.
       67. Van Horn, L., C. Ballew, K. Liu, K. Ruth, A. McDonald,
   J. E. Hilner, G. L. Burke, P. J. Savage, C. Bragg, B. Caan, D.
   Jacobs, M. Slattery, S. Sidney, ``Diet, Body Size, and Plasma
   Lipids-lipoproteins in Young Adults: Differences by Race and
   Sex,'' American Journal of Epidemiology, 133:9 - 22, 1991.
       68. Van Horn, L., Emidy, L. A., Liu, K., Liao, Y., Ballew,
   C., King, J., Stamler, J. 1988, Serum ``Lipid Response to a
   Fat-modified, Oatmeal-enhanced Diet,'' Preventive Medicine,
   17:377 - 386.
       69. Van Horn, L., A. Moag-Stahlberg, K. Liu, C. Ballew, K.
   Ruth, R. Hughes, J. Stamler, ``Effects on Serum Lipids of
   Adding Instant Oats to Usual American Diets,'' American
   Journal of Public Health, 81:183 - 88, 1991.
       70. Venter, C. S., H. H. Vorster, D. G. Van der Nest,
   ``Comparison Between Physiological Effects of
   Konjac-glucomannan and Propionate in Baboons Fed Western
   Diets,'' Journal of Nutrition, 120:1046 - 1053, 1991.
       71. World Health Organization, ``Diet, Nutrition, and the
   Prevention of Chronic Diseases,'' WHO, Geneva, Switzerland,
   1990.
       72. Nishina, P. M., R. A. Freedland, ``Effects of
   Propionate on Lipid Biosynthesis in Isolated Rat
   Hepatocytes,'' Journal of Nutrition, 120:668 - 673, 1990.
       73. Story, J. A., ``Modification of Steroid Excretion in
   Response to Dietary Fiber: Basic and Clinical Aspects,''
   Vahouny, G. V., D., Kritchevsky, Eds., New York: Plenum Press,
   pp. 253 - 264.
       74. Sugano, M., I. Ikeda, K. Imaizumi, Y. Lu, ``Dietary
   Fiber--
   Chemistry, Physiology, and Health Effects,'' Kritchevsky, D.,
   Bonfield, C., Anderson, J. W., eds., New York: Plenum Press,
   p. 137 - 156, 1990.
       75. Vahouny, G. V., M. M. Cassidy, ``Dietary Fibers and
   Absorption of Nutrients,'' Proceeding of the Society for
   Experimental Biology and Medicine, 180:432 - 446, 1985.
       76. National Cancer Institute, ``Cancer Prevention: Good
   News, Better News, Best News,'' NIH Publ. No. 84 - 2671.
   National Institutes of Health (NIH), Public Health Service
   (PHS), Department of Health and Human Services (DHHS), U.S.
   Government Printin
       77. National Cancer Institute, ``Diet, Nutrition & Cancer
   Prevention: to Food Choices,'' NIH Publ. No. 87 - 2878, NIH,
   PHS, DHHS, U.S. Government Printing Office, Washington, DC,
   1987.
       78. National Cancer Institute, ``Eat More Fruits and
   Vegetables* * * Five a Day for Better Health,'' NIH Pub. No.
   92 - 3248, NIH, PHS, DHHS, U.S. Government Printing Office,
   Washington, DC, 1991.
       79. Carroll, M., C. Sempos, R. Briefel, et al., ``Serum
   Lipids of Adults 20 - 74 years, United States, 1976 - 80,''
   Vital Health Statistics II, U.S. DHHS, HHS, National Center
   for Health Statistics, in press, 1992.
       80. DHHS, PHS and NIH, NCEP, ``Eating to Lower Your High
   Blood Cholesterol,'' NIH, Bethesda, MD, 1989.
       81. Institute of Medicine, National Academy of Sciences,
   ``Nutrition Labeling* * * Issues and Directions for the
   1990s,'' National Academy Press, Washington, DC, 1990.
       82. Burr, M. L., ``Vegetarianism, Dietary Fiber, and
   Mortality,'' American Journal of Nutrition, 36:873 - 877,
   1982.
       83. Leadbetter, J., M. J. Ball, J. I. Mann, ``Effects of
   Increasing Quantities of Oat Bran in Hypercholesterolemic
   People,'' American Journal of Clinical Nutrition, 54:841 -
   845, 1991.
       84. Cara, L., P. Borel, M. Armand, M. Senft, H. Lafont, et
   al., ``Plasma Lipid Lowering Effects of Wheat Germ in
   Hypercholesterolemic Subjects,'' Plant Foods for Human
   Nutrition, 41:135 - 150, 1991.
       85. Karlander, S., I. Armyr, S. Efendic, ``Metabolic
   Effects and Clinical Value of Beet Fiber Treatment in NIDDM
   Patients,'' Diabetes Research and Clinical Practice, 11:65 -
   72, 1991.
       86. Israelsson, B., G. Jarnblad, K. Persson, ``Serum
   Cholesterol Reduced with Fibrex, a Sugar-beet Fiber
   Preparation,'' unpublished, submitted with comments from the
   food industry, 1991.
       87. Spiller, G. A., J. W. Farquhar, J. E. Gates, S. F.
   Nichols, ``Guar Gum and Plasma Cholesterol, Effect of Guar Gum
   and an Oat Fiber Source on Plasma Lipoproteins and Cholesterol
   in Hypercholesterolemic Adults,'' Arteriosclerosis and
   Thrombosis, 11:1204 - 1208, 1991.
       88. Kawatra, A., A. C. Kapoor, S. Seghal,
   ``Hypocholesterolemic Effect of Guar Gum in Overweight
   Adults,'' Plant Foods for Human Nutrition, 41:241 - 245, 1991.
       89. Tinker, L. F., B. O'Schneeman, P. A. Davis, D. D.
   Gallaher, C. R. Waggoner, ``Consumption of Prunes as a Source
   of Dietary Fiber in Men with Mild Hypercholesterolemia,''
   American Journal of Clinical Nutrition, 53:1259 - 1265, 1991.
       90. Earll, J. M., ``Feasibility and Metabolic Effects of a
   Precisely Measurable Dietary Food Supplement,'' unpublished
   report submitted with comments, 1986.
       91. Bremer, J. M., R. S. Scott, C. J. Lintott, ``Oat Bran
   and Cholesterol Reduction: Evidence Against Specific Effect,''
   Australia and New Zealand Journal of Medicine, 21:422 - 426,
   1991.
       92. Anderson, J. W., T. L. Floore, P. B. Geil, D. S.
   O'Neal, T. K. Balm, ``Hypocholesterolemic Effects of Different
   Bulk-forming Hydrophilic Fibers as Adjuncts to Dietary Therapy
   in Mild to Moderate Hypercholesterolemia,'' Archives of
   Internal Medicine, 151:1597 - 1602, 1991.
       93. Jenkins, D. J. A., S. Mueller, T. M. S. Wolever, V.
   Rao, T. Ransom, D. Boctor, et al., ``High Soluble Fiber Foods
   Reduce Serum Lipids Even When Diets are Already Low in
   Saturated Fat and Cholesterol,'' unpublished report submitted
   from industry,1991.
       94. Jenkins, D. J. A., T. M. S. Wolever, V. Rao, R.
   Hegele, S. Mitchell, et al., ``High Fiber Foods Reduce Serum
   Lipids Even on Diets Low in Saturated Fat and Cholesterol,''
   unpublished report from industry, 1991.
       95. Anderson, J. W., ``Cholesterol-lowering Effects of
   Psyllium-enriched Cereal,'' unpublished report from industry,
   1991.
       96. Resnicow, K., J. Barone, A. Engle, S. Miller, N.
   Haley, et al., ``Diet and Serum Lipids in Vegan Vegetarians: A
   Model for Risk Reduction,'' Journal of the American Dietetic
   Association, 91:447 - 453, 1991.
       97. Kirsten, R., B. Domning, K. Nelson, K. Nemeth, G.
   Oremek, et al., ``Influence of Guar on Serum Lipids in
   Patients with Hyperlipidaemia,'' European Journal of Clinical
   Pharmacology, 37:117 - 120, 1989.
       98. Neal, G. W., T. K. Balm, ``Synergistic Effects of
   Psyllium in the Dietary Treatment of Hypercholesterolemia,''
   Southern Medical Journal, 83:1131 - 1137, 1990.
       99. Niemi, M. K., S. M. Keinanen-Kiukaanniemi, P. I.
   Salmela, ``Long-term Effects of Guar Gum and Microcrystalline
   Cellulose on Glycaemic Control and Serum Lipids in Type 2
   Diabetes,'' European Journal of Clinical Pharmacology, 34:427
   - 429, 1988.
       100. Uusitupa, M., O. Siitonen, K. Savolainen, M.
   Silvasti, I. Penttila, M. Parviainen, ``Metabolic and
   Nutritional Effects of Long-term Use of Guar Gum in the
   Treatment of Noninsulin-dependent Diabetes of Poor Metabolic
   Control,'' American Journal of Clinical Nutrition, 49:345 -
   351, 1989.
       101. Wolever, T., D. J. A. Jenkins, S. Mueller, C.
   Mehling, L. Katzman, et al., ``Psyllium Reduces Blood Lipids
   to a Similar Extent in Hyperlipidemic Men and Women,''
   unpublished study submitted with comments.
       102. Wolever, T., D. J. A. Jenkins, S. Mueller, D. Boctor,
   T. P. P. Ransom, et al., ``Method of Administration Influences
   the Serum Cholesterol-lowering Effect of Psyllium,''
   unpublished study submitted with comments.
       103. O'Connor, R. R., J. Jin, D. S. Hwang, ``A Multicenter
   Study of a Five Fiber Supplement in the Treatment of
   Hypercholesterolemia,'' final report, unpublished study
   submitted with comments, 1992.
       104. Whyte, J., R. McArthur, D. Topping, P. Nestel, ``Oat
   Bran Lowers Plasma Cholesterol in Mildly Hypercholesterolemic
   Men,'' Journal of the American Dietetic Association, 92:446 -
   449, 1992.
       105. Cerda, J. J., F. L. Robbins, C. W. Burgin, T. G.
   Baumgartner, R. W. Rice, ``The Effects of Grapefruit Pectin on
   Patients at Risk for Coronary Heart Disease Without Altering
   Diet or Lifestyle,'' Clinical Cardiology, 11:589 - 594, 1988.
       106. Haskell, W. L., G. A. Spiller, C. D. Jensen, B. K.
   Ellis, J. E. Gates, ``Role of Water-Soluble Dietary Fiber in
   the Management of Elevated Plasma Cholesterol in Healthy
   Subjects,'' The American Journal of Cardiology, 69:433 - 439,
   1992.
       107. Nervi, F., C. Covarrubias, P. Bravo, N. Velasco, N.
   Ulloa, et al., ``Influence of Legume Intake on Biliary Lipids
   and Cholesterol Saturation in Young Chilean Men,'' Journal of
   Clinical Nutrition and Gastroenterology, 96:825 - 830, 1989.
       108. Unpublished study submitted from food industry with
   comments, ``Cholesterol-lowering Effects of Ready-to-eat
   Cereal Containing Psyllium,'' 1991.
       109. Spencer, K. M., D. L. Gee, ``The Cholesterol Lowering
   Effect of a Fiber Supplemented Apple Juice Beverage in Men
   with Mild Hypercholesterolemia,'' unpublished study submitted
   with comments from food industry, 1991.
       110. Anderson, J. W., L. Story, B. Sieling, W. L. Chen,
   ``Hypocholesterolemic Effects of High Fibre Diets Rich in
   Water-soluble Plant Fibres,'' Journal of the Canadian Dietetic
   Association, vol. 45, No. 2, 1984.
       111. Turnbull, W. H., A. R. Leeds, ``Reduction of Total
   and LDL-cholesterol in Plasma by Rolled Oats,'' Journal of
   Clinical Nutrition and Gastroenterology, 2:1 - 4, 1987.
       112. Crane, N., memo to file, Clinical Nutrition Branch,
   October 15, 1992.
       113. Anderson, J. W., N. H. Gilinsky, D. A. Deakins, S. F.
   Smith, D. S. O'Neal, D. W. Dillon, P. R. Oeltgen, ``Lipid
   Responses of Hypercholesterolemic Men to Oat-bran and Wheat
   bran Intake,'' American Journal of Clinical Nutrition, 54:678
   - 683, 1991.
       114. Fukagawa, N. K., J. W. Anderson, G. Hageman, V. R.
   Young, K. L. Minaker, ``High-carbohydrate, High-fiber Diets
   Increase Peripheral Insulin Sensitivity in Healthy Young and
   Old Adults,'' American Journal of Clinical Nutrition, 52:524 -
   528, 1990.
       115. Reference letter from Kellogg Co. to FDA, December 5,
   1990.
       116. Reference letter from FDA to Kellogg Co., December
   12, 1990.
       117. Walsh, D. E., V. Yaghoubian, A. Behforooz, ``Effect
   of Glucomannan on Obese Patients: A Clinical Study,''
   International Journal of Obesity, 8:289 - 293, 1983.
       118. Anderson, J. W., S. Riddell-Mason, N. J. Gustafson,
   et al., ``Cholesterol-lowering Effects of Psyllium-enriched
   Cereal as an Adjunct to a Prudent Diet in the Treatment of
   Mild to Moderate Hypercholesterolemia,'' American Journal of
   Clinical Nutrition, 56:93 - 98, 1992.
       119. Braaten, J. T., M. K. L. Braaten, P. Cradley-White,
   et al., ``Cholesterol Lowering Effect of Oat Gum in
   Hypercholesterolemic Subjects,'' unpublished study, 1992.
       120. Bridges, S. R., J. W. Anderson, D. A. Deakins et al.,
   ``Oat Bran Increases Serum Acetate of Hypercholesterolemic
   Men,''American Journal of Clinical Nutrition, 56:455 - 459.
       121. Kashtan, H. H., S. Stern, D. J. A. Jenkins, et al.,
   ``Wheat bran and Oat-bran Supplements' Effects on Blood Lipids
   and Lipoproteins,'' American Journal of Clinical Nutrition,
   55:976 - 980, 1992.
       122. Ranhotra, G. S., J. A. Gelroth, R. D. Reeves, et al.,
   ``Short-term Lipidemic Responses in Otherwise Healthy
   Hypercholesterolemic Men Consuming Foods High in Soluble
   Fiber,'' Cereal Chemistry, 66:94 - 97, 1989.
       123. Zhang, J. G. Hallmans, H. Andersson, et al., ``Effect
   of Oat Bran on Plasma Cholesterol and Bile Acid Excretion in
   Nine Subjects with Ileostomies,'' American Journal of Clinical
   Nutrition, 56:99 - 105, 1992.
       124. Marlett, J. A., K. B. Hosig, N. W. Vollendorf, et
   al., ``Mechanism or Serum Cholesterol Reduction by Oat Bran,''
   unpublished study submitted with comments, 1992.
       125. Ripsin, C. M., J. M. Keenan, D. R. Jacobs, et al.,
   ``Oat Products and Lipid Lowering--A Meta-analysis,'' Journal
   of the American Medical Association, 267:3317 - 3325, 1992.
       126. Turnbull, W. H., A. R. Leeds, ``The Effect of Rolled
   Oats and a Reduced/Modified Fat Diet on Apolipoprotein AI and
   B,'' Journal of Clinical and Nutritional Gastroenterology,
   1:15 - 19, 1989.
       127. Ney, D. M., J. B. Lasekan, F. L. Shinnick, ``Soluble
   Oat Fiber Tends to Normalize Lipoprotein Composition in
   Cholesterol-fed Rats,'' Journal of Nutrition, 118:1455 - 1462,
   1988.
       128. Nishina, P. M., B. O. Schneeman, R. A. Freedland,
   ``Effects of Dietary Fibers on Nonfastin Plasma Lipoprotein
   and Apolipoprotein Levels in Rats,'' Journal of Nutrition,
   121:431 - 437, 1991.
       129. Prentice, N., A. A. Qureshi, W. C. Burger, C. E.
   Elson, ``Response of Hepatic Cholesterol, Fatty Acid Synthesis
   and Activities of Related Enzymes to Rolled Barley and Oats in
   Chickens,'' Nutrition Reports International, 26:597 - 604,
   1982.
       130. O'Mahony, S., ``Biofunctional Properties of
   Oat-cereal Fractions. No. 7,'' unpublished studies submitted
   with comments, 1975.
       131. Qureshi, A. A., W. C. Burger, N. Prentice, et al.,
   ``Regulation of Lipid Metabolism in Chicken Liver by Dietary
   Cereals,'' Journal of Nutrition, 110:388 - 393, 1980.
       132. Ranhorta, G. S., J. A. Gelroth, K. Astroth, C. S.
   Rao, ``Relative Lipidemic Responses in Rats Fed Oat Bran or
   Oat Bran Concentrate,'' Cereal Chemistry, 67:509 - 511, 1990.
       133. Shinnick, F. L., M. J. Longacre, S. L. Ink, J. A.
   Marlett, ``Oat Fiber: Composition Versus Physiological
   Function in Rats,'' Journal of Nutrition, 118:144 - 151, 1988.
       134. Shinnick, F. L., S. L. Ink, J. A. Marlett,
   ``Dose-Response to a Dietary Oat Bran Fraction in
   Cholesterol-Fed Rats,'' Journal of Nutrition, 120:561 - 568,
   1990.
       135. Rogel, A. M., P. Vohra, ``Alteration of Lipid
   Metabolism in Japanese Quail by Feeding Oat Hulls and Brans,''
   Poultry Science, 62: 1045 - 1053, 1983.
       136. O'Mahony, S., ``Biofunctional Properties of
   Oat-cereal Fractions. No. 18,'' unpublished studies submitted
   with comments, 1976.
       137. Bengtsson, S., P. Aman, H. Graham, et al., ``Chemical
   Studies on Mixed-Linked [-Glucans in Hull-less Barley
   Cultivars Giving Different Hypocholesterolemic Responses in
   Chickens,'' Journal of Science and Food Agriculture, 52:435 -
   445,
       138. Kritchevsky, D., S. A. Tepper, G. T. Goodman, et al.,
   ``Influence of Oat and Wheat Bran on Cholesterolemia in
   Rats,'' Nutrition Reports International, 29:1353 - 1359, 1984.
       139. Klopfenstein, C. F., R. C. Hoseney, ``Cholesterol
   Lowering Effect of Beta-Glucan-Enriched Bread,'' Nutrition
   Reports International, 36:1091 - 1098, 1987.
       140. Kahlon, T. S., R. M. Saunders, F. I. Chow, et al.,
   ``Influence of Rice Bran, Oat Bran, and Wheat Bran on
   Cholesterol and Triglycerides in Hamsters,'' Cereal Chemistry,
   67:439 - 443, 1990.
       141. Jennings, C. D., K. Boleyn, S. R. Bridges, et al.,
   ``A Comparison of the Lipid-Lowering and Intestinal
   Morphological Effects of Cholestyramine, Chitosan, and Oat Gum
   in Rats (42773),'' Proceedings of the Society for Experimental
   Biology and Medicine, 189:13 - 20, 1988.
       142. Forsythe, W. A., W. L. Chenoweth, M. R. Bennink,
   ``Laxation and Serum Cholesterol in Rats Fed Plant Fibers,''
   Journal of Food Science, 43:1470 - 1472, 1978.
       143. Chen, W. L., J. W. Anderson, M. R. Gould, ``Effects
   of Oat Bran, Oat Gum and Pectin on Lipid Metabolism of
   Cholesterol-fed Rats,'' Nutrition Reports International,
   24:1093 - 1098, 1981.
       144. Chen, W. L., J. W. Anderson, ``Effects of Plant Fiber
   in Decreasing Plasma Total Cholesterol and Increasing
   High-Density Lipoprotein Cholesterol (40671),'' Proceedings of
   the Society for Experimental Biology and Medicine, 162:310 -
   313, 1979.
       145. Fadel, J. G., R. K. Newman, C. W. Newman, et al.,
   ``Hypocholesterolemic Effects of Beta-Glucans in Different
   Barley Diets Fed to Broiler Chicks,'' Nutrition Reports
   International, 35:1049 - 1058, 1987.
       146. Chen, W. L., J. W. Anderson, ``Effects of Guar Gum
   and Wheat Bran on Lipid Metabolism of Rats,'' Journal of
   Nutrition, 109:1028 - 1034, 1979.
       147. Turley, S. D., B. P. Daggy, J. M. Dietschy,
   ``Cholesterol-lowering Action of Psyllium Mucilloid in the
   Hamster: Sites and Possible Mechanisms of Action,''
   Metabolism, 40:1063 - 1073, 1991.
       148. Wilson, J. N., S. P. Wilson, R. P. Eaton, ``Dietary
   Fiber and Lipoprotein Metabolism in the Genetically Obese
   Zucker Rat,'' Arteriosclerosis, 4:147 - 153, 1984.
       149. Welch, R. W., E. M. Peterson, B. Schramka,
   ``Hypocholesterolemic and Gastrointestinal Effects of Oat Bran
   Fractions in Chicks,'' Nutrition Reports International, 38:551
   - 561, 1988.
       150. Hundemer, J. K., S. P. Nabar, B. J. Shriver, L. P.
   Forman, ``Dietary Sources Lower Blood Cholesterol in C57BL/6
   Mice,'' Journal of Nutrition, 121:1360 - 65, 1991.
       151. Ginker, E., F. J. Kubec, J. Vozar, P. Bobek,
   ``Natural Hypocholesterolemic Agent: Pectin Plus Ascorbic
   Acid,'' International Journal of Vitamin Research, 49:406 -
   412, 1979.
       152. Kakis, G., E. Chao, M. Schmidt, et al., ``Psyllium
   Induced Cholesterol Lowering in the Hamster,'' unpublished
   study submitted with comments, 1992.
       153. LSRO, FASEB, ``Evaluation of the Health Aspects of
   Oat Gum, Okra Gum, Quince Gum, and Psyllium Husk Gum as Food
   Ingredients,'' Bethesda, MD, 1982.
       154. Schrijver, R. D., D. Fremaut, A. Verheyen,
   ``Cholesterol-lowering Effects and Utilization of Protein,
   Lipid, Fiber and Energy in Rats Fed Unprocessed and Baked Oat
   Bran,'' Journal of Nutrition, 122:1318-1324, 1992.
       155. Lenfant, C., Letter to David Kessler, October 29,
   1992.
   
   List of Subjects in 21 CFR Part 101
   
       Food labeling, Reporting and recordkeeping requirements.
       Therefore, under the Federal Food, Drug, and Cosmetic Act,
   and under authority delegated to the Commissioner of Food and
   Drugs, 21 CFR part 101 is amended as follows:
   
   
   PART 101--FOOD LABELING
   
       1. The authority citation for 21 CFR part 101 continues to
   read as follows:
   
       Authority: Secs. 4, 5, 6 of the Fair Packaging and
   Labeling Act (15 U.S.C. 1453, 1454, 1455); secs. 201, 301,
   402, 403, 409, 701 of the Federal Food, Drug, and Cosmetic Act
   (21 U.S.C. 321, 331, 342, 343, 348, 371).
   
       2. New 101.71 is added to subpart E to read as follows:
   
   
   101.71  Health claims: claims not authorized.
   
   * * * * *
       (b) Dietary fiber and cardiovascular disease.
   
       3. New 101.77 is added to subpart E to read as follows:
   
   
   101.77  Health claims: fruits, vegetables, and grain products
   that contain fiber, particularly soluble fiber, and risk of
   coronary heart disease.
   
       (a) Relationship between diets low in saturated fat and
   cholesterol and high in fruits, vegetables, and grain products
   that contain fiber, particularly soluble fiber, and risk of
   coronary heart disease. (1) Cardiovascular disease means
   diseases of the heart and circulatory system. Coronary heart
   disease is the most common and serious form of cardiovascular
   disease and refers to diseases of the heart muscle and
   supporting blood vessels. High blood total- and low density
   lipoprotein (LDL)- cholesterol levels are major modifiable
   risk factors in the development of coronary heart disease.
   High coronary heart disease rates occur among people with high
   blood cholesterol levels of 240 milligrams per deciliter
   (mg/dL) (6.21 (mmol/L)) or above and LDL-cholesterol levels of
   160 mg/dL (4.13 mmol/L) or above. Borderline high risk blood
   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. Dietary lipids (fats) include fatty acids and
   cholesterol. Total fat, commonly referred to as fat, is
   composed of saturated fat (fatty acids containing no double
   bonds), and monounsaturated and polyunsaturated fat (fatty
   acids containing one or more double bonds).
       (2) 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 coronary heart disease. Diets low in
   saturated fat and cholesterol are associated with decreased
   levels of blood total- and LDL-cholesterol, and thus, with
   decreased risk of developing coronary heart disease.
       (3) Populations with relatively low blood cholesterol
   levels tend to have dietary patterns that are not only low in
   total fat, especially saturated fat and cholesterol, but are
   also relatively high in fruits, vegetables, and grain
   products. Althoughthe specific roles of these plant foods are
   not yet fully understood, many studies have shown that diets
   high in plant foods are associated with reduced risk of
   coronary heart disease. These studies correlate diets rich in
   fruits, vegetables, and grain products and nutrients from
   these diets, such as some types of fiber, with reduced
   coronary heart disease risk. Persons consuming these diets
   frequently have high intakes of dietary fiber, particularly
   soluble fibers. Currently, there is not scientific agreement
   as to whether a particular type of soluble fiber is
   beneficial, or whether the observed protective effects of
   fruits, vegetables, and grain products against heart disease
   are due to other components, or a combination of components,
   in these diets, including, but not necessarily limited to,
   some types of soluble fiber, other fiber components, other
   characteristics of the complex carbohydrate content of these
   foods, other nutrients in these foods, or displacement of
   saturated fat and cholesterol from the diet.
       (b) Significance of the relationship between diets low in
   saturated fat and cholesterol, and high in fruits, vegetables,
   and grain products that contain fiber, particularly soluble
   fiber, and risk of coronary heart disease. (1) Coronary heart
   disease is a major public health concern in the United States,
   primarily because it accounts for more deaths than any other
   disease or group of diseases. Early management of risk factors
   for coronary heart disease is a major public health goal that
   can assist in reducing risk of coronary heart disease. There
   is a continuum of mortality risk from coronary heart disease
   that increases with increasing levels of blood
   LDL-cholesterol. Individuals with high blood LDL-cholesterol
   are at greatest risk. A largers that reducing saturated fat
   and cholesterol intakes lowers blood LDL-cholesterol and risk
   of heart disease in most individuals, including persons with
   blood cholesterol levels in the normal range. Additionally,
   consuming diets high in fruits, vegetables, and grain
   products, foods that contain soluble fiber, may be a useful
   adjunct to a low saturated fat and low cholesterol diet.
       (2) Other risk factors for coronary heart disease include
   a family history of heart disease, high blood pressure,
   diabetes, cigarette smoking, obesity (body weight 30 percent
   greater than ideal body weight), and lack of regular physical
   exercise.
       (3) Intakes of saturated fat exceed recommended levels in
   many people in the United States. Intakes of cholesterol are,
   on average, at or above recommended levels. Intakes of
   fiber-containing fruits, vegetables, and grain products are
   about half of recommended intake levels. One of the major
   public health recommendations relative to coronary heart
   disease 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. Recommended total dietary
   fiber intakes are about 25 grams (g) daily, of which about 25
   percent (about 6 g) should be soluble fiber.
       (4) Current dietary guidance recommendations encourage
   decreased consumption of dietary fat, especially saturated fat
   and cholesterol, and increased consumption of fiber-rich foods
   to help lower blood LDL-cholesterol levels. Results of
   numerous studies have shown that fiber-containing fruits,
   vegetables, and grain products can help lower blood
   LDL-cholesterol.
       (c) Requirements. (1) All requirements set forth in
   101.14 shall be met.
       (2) Specific requirements. (i) Nature of the claim. A
   health claim associating diets low in saturated fat and
   cholesterol and high in fruits, vegetables, and grain products
   that contain fiber, particularly soluble fiber, with reduced
   risk of heart disease may be made on the label or labeling of
   a food described in paragraph (c)(2)(ii) of this section,
   provided that:
       (A) The claim states that diets low in saturated fat and
   cholesterol and high in fruits, vegetables, and grain products
   that contain fiber ``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) The claim is limited to those fruits, vegetables, and
   grains that contain fiber;
       (D) In specifying the dietary fiber, the claim uses the
   term ``fiber,'' ``dietary fiber,'' ``some types of dietary
   fiber,'' ``some dietary fibers,'' or ``some fibers;'' the term
   ``soluble fiber'' may be used in addition to these terms;
       (E) In specifying the fat component, the claims uses the
   terms ``saturated fat'' and ``cholesterol;'' and
       (F) The claim indicates that development of heart disease
   depends on many factors; and
       (G) The claim does not attribute any degree of risk
   reduction for coronary heart disease to diets low in saturated
   fat and cholesterol and high in fruits, vegetables, and grain
   products that contain fiber.
       (ii) Nature of the food. (A) The food shall be or shall
   contain a fruit, vegetable, or grain product.
       (B) The food shall meet the nutrient content requirements
   of 101.62 for a ``low saturated fat,'' ``low cholesterol,''
   and ``low fat'' food.
       (C) The food contains, without fortification, at least 0.6
   g of soluble fiber per reference amount customarily consumed;
       (D) The content of soluble fiber shall be declared in the
   nutrition information panel, consistent with
   101.9(c)(6)(i)(A).
       (d) Optional information. (1) The claim may identify one
   or more of the following risk factors for heart disease about
   which there is general scientific agreement: A family history
   of coronary heart disease, elevated blood-, total- and
   LDL-chole, and physical inactivity.
       (2) The claim may indicate that the relationship of diets
   low in saturated fat and cholesterol, and high in fruits,
   vegetables, and grain products that contain fiber to 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 low in saturated fat and cholesterol and high in
   fruits, vegetables, and grain products that contain fiber and
   coronaryheart disease, and the significance of the
   relationship.
       (4) In specifying the nutrients, the claim may include the
   term ``total fat'' in addition to the terms ``saturated fat''
   and ``cholesterol.''
       (5) The claim may indicate that it 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).
       (6) 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.
       (7) 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 claims. The following model health claims
   may be used in food labeling to characterize the relationship
   between diets low in saturated fat and cholesterol and high in
   fruits, vegetables, and grain products that contain
   solublefiber:
       (1) Diets low in saturated fat and cholesterol and rich in
   fruits, vegetables, and grain products that contain some types
   of dietary fiber, particularly soluble fiber, may reduce the
   risk of heart disease, a disease associated with many factors.
       (2) Development of heart disease depends on many factors.
   Eating a diet low in saturated fat and cholesterol and high in
   fruits, vegetables, and grain products that contain fiber may
   lower blood cholesterol levels and reduce your risk of heart
   disease.
   
       Dated: November 6, 1992.
   David A. Kessler,
   Commissioner of Food and Drugs.
   Louis W. Sullivan,
   Secretary of Health and Human Services.
       Note: The following table will not appear in the annual
   Code of Federal Regulations.
   
   BILLING CODE 4160 - 01 - F
   
   
   [FR Doc. 92 - 31513 Filed 12 - 28 - 92; 8:45 am]
   BILLING CODE 4160 - 01 - C