• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

About FDA

  • Print
  • Share
  • E-mail

FDA Response to General Mills Addressing the Scientific Basis for Cheerios Cholesterol Claims

October 9, 2009

Janice Marturano
Vice President
Public Responsibility & Deputy General Counsel
General Mills
One General Mills Boulevard
Minneapolis, Minnesota 55426

Dear Ms. Marturano:

We have conducted a preliminary review of the studies you submitted to us during our June 1, 2009 meeting. These studies were provided as support for "4% in 6 weeks" and "10% in one month" LDL cholesterol reduction claims that have appeared on the label of your Cheerios™ cereal in connection with a health claim for soluble fiber and coronary heart disease (see 21 C.F.R. 101.81). The preliminary comments in this letter are limited to the studies you submitted to us and to the scientific evidence discussed in the 1996-97 rulemaking to authorize the health claim on soluble fiber from whole oats and reduced risk of coronary heart disease. At this time, we are not addressing the revised labels that were attached to your July 14, 2009 e-mail or your legal arguments regarding whether your "4% in 6 weeks" or "10% in one month" LDL cholesterol reduction claims are permitted in accordance with the current regulations governing the soluble fiber and coronary heart disease health claim. We will address these legal arguments and the revised labels in separate correspondence.

Before we comment on the studies you submitted, some context on the bigger picture is in order. By law, FDA's health claim evaluations are based on the totality of the publicly available scientific evidence (see 21 U.S.C. § 343(r)(3)(B)(i)). For a health claim to include a statement about magnitude of effect, that statement -- like any other statement in a health claim -- must be supported by the totality of the evidence. The studies you provided to us in support of your claims are not the only studies evaluating the cholesterol-lowering effects of foods containing soluble fiber from whole oats that have been published since the issuance of the 1997 final rule authorizing the soluble fiber/coronary heart disease health claim. To decide whether the total body of publicly available scientific evidence supports amending the soluble fiber/coronary heart disease health claim to provide for an optional statement that soluble fiber from whole oats can reduce LDL cholesterol by a specific percentage, we would have to review all relevant studies that have evaluated the relationship between soluble fiber from whole oats and LDL cholesterol reduction for a sufficient duration, not just the Cheerios studies you submitted.

Four studies submitted by General Mills evaluated the effect of consuming soluble fiber from whole oats on blood cholesterol levels. These studies all used Cheerios as the dietary source of soluble fiber from whole oats. You cited three of these studies as support for the "4% in 6 weeks" LDL cholesterol reduction claims that were the subject of FDA's May 5, 2009, warning letter to General Mills (Johnston et al., 1998; Karmally et al., 2005; Reynolds at al., 2000), and one study as support for a "10% in one month" LDL cholesterol reduction claim (Maki et al., 2009) that appeared on another version of the Cheerios label. Our comments on these studies follow.

Studies Submitted to Support Claim for 4% LDL Cholesterol Reduction in Six Weeks

1. Johnston et al. (1998)

Description of Study: Johnston et al. (1998) was a double-blind, randomized, parallel-design study that included 120 men and women 40 to 70 years of age with mild to moderate hypercholesterolemia. The subjects consumed a Step 1 diet (total fat ≤ 30% of total calories, saturated fat ≤ 10% of total calories, cholesterol < 300 mg/day) and either 3 g/day of soluble fiber from Cheerios, which contains whole oats, or an essentially fiber-free (0.1 g/day) cereal as a control. After 6 weeks on these diets, it was reported that there was a significant reduction in LDL cholesterol (- 4.2%) from consuming Cheerios as compared to the control cereal.

Comments: Based on our preliminary review, this study appears to provide data that soluble fiber from whole oats can lower LDL cholesterol by an average of 4% in six weeks.

2. Karmally et al. (2005)

Description of Study: Karmally et al. (2005) was a randomized parallel study that included approximately 145 Hispanic men and women 30 to 70 years of age with normal or high blood cholesterol levels. The subjects consumed a Step 1 diet and either 3 g/day of soluble fiber from Cheerios, which contains whole oats, or a soluble fiber-free corn cereal as a control. After 6 weeks on the Cheerios diet, it was reported that the average reduction in LDL cholesterol, compared to baseline, was -5.3%. The average change (pre- versus post-treatment) in the subjects' blood concentration of total and LDL cholesterol (mg/dL) was provided for each of the two groups (control and Cheerios), as well as the average percent reduction in cholesterol levels for each group. The study did not, however, report the average change in total and LDL cholesterol levels in the Cheerios group compared to control (i.e., the net change in blood cholesterol levels), either in terms of mg/dL or as a percentage.

Comments: Based on our preliminary review, it is not possible to determine whether this study supports a claim that soluble fiber from whole oats can lower LDL cholesterol by an average of 4% in six weeks because the article did not report the average net percentage reduction in LDL cholesterol level in the Cheerios group as compared to the control group. Moreover, it is not possible to calculate the average net percentage reduction in LDL cholesterol from consuming Cheerios based on the information provided in the article. Therefore, it is unclear whether this study supports a "4% in 6 weeks" claim about LDL cholesterol reduction from consuming Cheerios.

During our June 12, 2009 meeting you presented a slide (slide #12) stating that the Karmally study demonstrates a 7.5% reduction in LDL cholesterol. According to slide #13 in your presentation, the 7.5% reduction represents the change in LDL cholesterol relative to the control. However, it is not clear how this 7.5% reduction was calculated.

3. Reynolds et al. (2000)

Description of Study: Reynolds et al. (2000) was a double blind, randomized parallel-design study that included a total of 43 men and women 27 to 68 years of age with mild to moderate hypercholesterolemia. The subjects consumed a Step 1 diet and either 2.7 g/day of soluble fiber from Cheerios or an essentially fiber-free (0.2 g/day) cereal as a control. After 4 weeks on these diets, it was reported that the average reduction in LDL cholesterol from consuming Cheerios as compared to the control cereal was 4.9%.

Comments: Based on our preliminary review, this study would not support a claim that soluble fiber from whole oats can lower LDL cholesterol by an average of 4% in 6 weeks. The study was conducted for 4 weeks, and therefore does not provide data on the reduction of LDL cholesterol at 6 weeks. It cannot be determined from this study whether the consumption of Cheerios for 6 weeks would sustain the same magnitude of reduction in LDL cholesterol as when the cereal is consumed for 4 weeks.

Other Issues

In your May 14, 2009 letter to Tyra Wisecup, FDA Minneapolis District Compliance Officer, and during our June 12, 2009 meeting, you stated that the three studies submitted to support a claim for a 4% LDL cholesterol reduction in six weeks are consistent with many studies that were cited by FDA in promulgating the soluble fiber/CHD health claim. To support this assertion, you presented a slide (slide #13) during the June 12, 2009 meeting with a graph charting the percent change in LDL cholesterol relative to control for the FDA-reviewed studies1 and the three Cheerios studies. We reviewed the study descriptions in the preamble to the 1996 proposed rule for the health claim on soluble fiber from oats and coronary heart disease. See 61 Fed. Reg. 296 (Jan. 4, 1996). For the studies meeting the criteria cited in slide #13, the net change (treatment compared to control) in LDL cholesterol ranged from +1.4 to -16.9%.

A 4% reduction in LDL cholesterol was within the range reported from the studies reviewed in the proposed rule. However, these studies do not represent the totality of publicly available scientific evidence, because they do not include any studies published since the issuance of the 1997 final rule. Additionally, as described in the preamble to the 1996 proposed rule, these studies varied in duration, baseline cholesterol levels, and dose of soluble fiber. Therefore, it is not possible to rely on these studies to draw conclusions about the average reduction of LDL cholesterol from consuming 3 grams of soluble fiber from whole oats daily for six weeks.

In your July 9, 2009 letter to Michael Landa, you remark that FDA's own scientific conclusion, reached at the time it adopted the soluble fiber/coronary heart disease health claim regulation, is consistent with the "4% in 6 weeks" claim on the package of Cheerios. You cite as evidence the following statement from the 1997 final rule authorizing the soluble fiber/coronary heart disease health claim: "[A]n estimated intake of 3 g per day soluble fiber (i.e., β-glucan soluble fiber) is associated with a reduction in blood total cholesterol of about 5 percent." 62 Fed. Reg. 3584, 3589 (Jan. 23, 1997).

The quoted statement is part of FDA's discussion of a comment requesting consideration of a lower level of β-glucan soluble fiber per day (2.5 grams per day, rather than the 3 grams per day that FDA had proposed) as useful in reducing the risk of coronary heart disease. As background for FDA's response rejecting the comment, the agency described how the petition for the soluble fiber/coronary heart disease health claim had presented data from a dose-controlled study using a linear regression model to show the change in blood total cholesterol as a function of soluble fiber intake. The statement quoted in your letter is not, as you suggest, a statement of FDA's scientific conclusions based on the agency's own analysis of the totality of the publicly available evidence. Rather, it summarizes the results of the linear regression analysis in the petition, which relied on data from a single study. Thus, it is a description of an analysis presented in the health claim petition, not a statement of the agency's own views. FDA's evaluation of the evidence in its response to the comment describes the agency's conclusions in terms of levels of β-glucan and effectiveness in reducing the risk of coronary heart disease (e.g., "The results of the studies ... demonstrate that intakes of 3 g or more of β-glucan are more likely to be effective."; "[W]ithout further data, there is no justification for concluding that 2.5g per day is a more appropriate estimate of the amount of β-glucan useful in reducing the risk of coronary heart disease than is 3g per day."), not in terms of a specific level of reduction in blood total cholesterol. Similarly, throughout the final rule, FDA did not discuss its conclusions regarding the effects of soluble fiber in terms of a specific reduction in levels of blood cholesterol. Rather, the agency consistently described the benefits of soluble fiber using phrases such as "positive effects" on blood cholesterol levels or "effective in lowering serum lipids". In any event, your "4% in 6 weeks" claim is for "bad" (LDL) cholesterol, not total blood cholesterol, which is the subject of the statement you cited in your letter to Mr. Landa.

Additionally, because the magnitude of effect of soluble fiber on blood cholesterol levels can be affected by a number of factors (e.g., dose, control diet, population), FDA would need to review the total body of evidence to determine which factors need to be specified along with the claim to make it truthful and non-misleading. For example, the "4% in 6 weeks" and "10% in one month" claims are based on consumption of three servings of Cheerios per day. The health claim for soluble fiber and coronary heart disease requires disclosure of the amount of soluble fiber that is contributed by one serving of the product. This raises concerns of how percentage cholesterol reduction claims could be phrased to clearly communicate that the results are based on consumption of 3 grams of soluble fiber from oats daily, rather than a single serving of the product.

Study Submitted to Support Claim for 10% LDL Cholesterol Reduction in One Month

Description of Study: Maki et al. (2009) was a randomized, parallel-design study that included a total of 173 men and women 20 to 65 years of age with moderate hypercholesterolemia. The subjects were instructed to consume a weight reduction diet and either 3 g/day of soluble fiber from Cheerios or low-fiber breakfast/snack foods as a control. This study was conducted for 12 weeks and blood samples were collected at 4, 8, 10 and 12 weeks post-intervention. Both groups reduced their energy intake during the study and lost weight. There was no significant difference in weight loss between the two dietary groups. At 4 weeks (1 month), the change in LDL cholesterol levels from baseline was -10.7% for subjects consuming Cheerios and -6.2% for controls. When plasma levels were averaged for the 4, 8, 10 and 12 week collection periods and compared to baseline levels, the reduction in LDL cholesterol was -8.7% for the Cheerios group and -4.3% for the control group.

Comments: Based on our preliminary review, it is not possible to determine whether this study supports a claim that soluble fiber from whole oats can lower LDL cholesterol by an average of 10% in one month. The article did not report the average net percentage reduction in LDL cholesterol for the Cheerios group as compared to the control group, and it is not possible to accurately determine the average net percentage reduction based on the information provided in the article. In your slide presentation at our meeting on June 12, 2009, you cited to the 10.7% LDL cholesterol reduction from baseline for the Cheerios group at 4 weeks as the basis for the "10% in 1 month" claim. This 10.7% reduction represents the change in the subjects' LDL cholesterol levels since the beginning of the study and does not take into account the LDL cholesterol reduction observed in the control group (- 6.2%). If findings from the control group are not considered, it is not possible to ascertain whether changes in LDL cholesterol levels were due to the soluble fiber in Cheerios or due to other factors. For example, subjects in both groups consumed a weight reduction diet that resulted in weight loss, which is known to lower plasma cholesterol levels. Moreover, because the plasma levels were averaged for the 4, 8, 10, and 12 week collection periods and compared to baseline levels, rather than control, it is not possible to determine whether the 10% reduction at 4 weeks was sustained during the rest of the study and whether it reflects the final outcome of the study at 12 weeks. Thus, the findings reported in the study are not sufficient to support your "10% in 1 month" LDL cholesterol reduction claim for Cheerios.

In conclusion, in determining whether to amend 21 C.F.R. § 101.81 to authorize the soluble fiber/coronary heart disease claim to include a statement about a specific percentage reduction in LDL cholesterol, FDA would need to consider the totality of the publicly available scientific evidence to support such a statement and also how to convey information in a way that is not misleading.

If you have any questions please feel free to contact me at (301) 436-2373.

Sincerely,

Barbara O. Schneeman, Ph.D.
Director
Office of Nutrition, Labeling and Dietary Supplements
Center for Food Safety and Applied Nutrition

cc: Roderick A. Palmore
Executive Vice President
General Counsel & Chief Compliance and Risk Management Officer
General Mills, Inc.


1 The FDA studies graphed on the slide were chosen to meet the following criteria: intervention at least 3 weeks; provided up to 6.1 g oat beta-glucan; provided LDL concentration relative to control.