Did you hear the one about the FDA declaring Cheerios a drug?  Seems like everyone, including my close, personal, confirmed Facebook friends in the so-called Skeptical Movement, is teasing the FDA about the warning letter sent to General Mills threatening to consider Cheerios a drug.  And, to be honest, on the face of things it does seem a bit ridiculous; but, that is ok, we like ridiculous.  If you don't, boy you wound up in the wrong place.

Let's take a look at what the letter¹ actually says.  It describes the claims being made in the Cheerios advertising:

Specifically, your Cheerios® product bears the following claims ort [sic] its label:
• "you can Lower Your Cholesterol 4% in 6 weeks"
• "Did you know that in just 6 weeks Cheerios can reduce bad cholesterol by an average of 4 percent? Cheerios is ... clinically proven to lower cholesterol. A clinical study showed that eating two 1 1/2 cup servings daily of Cheerios cereal reduced bad cholesterol when eaten as part of a diet low in saturated fat and cholesterol."

The FDA lays out how these claims require Cheerios to be regulated as a drug:

These claims indicate that Cheerios® is intended for use in lowering cholesterol, and therefore in preventing, mitigating, and treating the disease hypercholesterolemia. Additionally, the claims indicate that Cheerios® is intended for use in the treatment, mitigation, and prevention of coronary heart disease through, lowering total and "bad" (LDL) cholesterol. . . Because of these intended uses, the product is a drug within the meaning of section 201(g)(1)(B) of the Act [21 U.S.C. § 321 (g)P)(B)]. The product is also a new drug under section 201(p) of the Act [21 U.S.C. § 321(p)] because it is not generally recognized as safe and effective for use in preventing or treating hypercholesterolemia or coronary heart disease. Therefore,under section 505(a) of the Act [21 U.S.C. § 355(a)], it may not be legally marketed with the above claims in the United States without an approved new drug application.

I know, it's freaking Cheerios; but, I say HUZZAH! for the FDA.  Sure, Cheerios is clearly a food, but what separates a food from a drug.  Food is made of chemicals.  Drugs are made of chemicals.  Food is eaten.  Some drugs are eaten.  Critically, under the law of the United States, a drug is a product that is marketed with a specific health claim, which requires FDA approval.  In 1997, the FDA approved the use of the "soluble fiber" claim (soluble fiber in oats can reduce cholesterol, thus reducing risk of heart disease²) for Cheerios marketing. 

If you still think considering Cheerios a drug is ridiculous, let's consider a hypothetical example.  Imagine that the American Carrot Farmers Association³ decides to advertise that eating carrots can reduce myopia by 10%.  If the ACFA is badly misrepresenting the research backing up the claim, they could wind up bitch slapped by the FTC like Kellogs did over their Frosted Mini-Wheats (Festival of Idiots #4) ad campaign.  What happens if the ACFA has funded clinical studies that have demonstrated that carrots improve eyesight?  The FTC does not have the health expertise to evaluate whether carrots should be used as a treatment for near-sightedness.  Should carrots be used instead of or in conjunction with traditional treatments like glasses and poking a finger in your eye (i.e., contacts)?  The only regulatory body that has the mandate and expertise to make such an evaluation is the FDA. 

If General Mills wishes to advertise Cheerios as a treatment for high cholesterol, then Cheerios must meet the same standards of both evidence and regulation as any other cholesterol treatment making similar claims.

So, how does the research look?  I found three different studies that General Mills has funded.  All three had similar results.  One study was published in a peer reviewed journal (Nutrition in Clinical Care), but that is no longer available even from the publisher's site.  The other two were reported at conferences.  Therefore, for descriptions of the studies I have had to rely on second-hand reports^2,4,5.

All three studies consistently showed 4-10% improvements in cholesterol levels.  Describing one of these studies, General Mills researcher⁶ Dr. Eric Gugger said: 

This is a clinical study-the gold standard for research-and it's
nice to
know that an American favorite such as Cheerios can be effective in
helping
women fight heart disease.

If Cheerios is meeting the GOLD STANDARD for research, they should not have any trouble sailing through FDA approval for over-the-counter sales.  Does the research design really meet the GOLD STANDARD for clinical research?  Not to nit-pick, but Dr. Gugger is wrong to assert that a clinical study is the GOLD STANDARD for biomedical research.  The GOLD STANDARD for biomedical research is the double-blind placebo-controlled⁷ clinical study.  

We are going to look at several criteria to see if these studies meet the GOLD STANDARD: is the treatment program reasonable, were the studies placebo-controlled, and were the studies double-blind?

Is the treatment reasonable?
According to previous studies, the beneficial cholesterol effects require 3g of dietary soluble fiber per day.  Cheerios contains 1g of soluble fiber per cup.  As a result, study participants in the experimental group had to eat 3 cups of Cheerios every day (~20% of total caloric intake) for 6 weeks.  The studies reported excellent compliance with the diet, although I don't think I could deal with it.

Were the studies placebo-controlled?
The purpose of a placebo is to make it impossible for individuals in the control group to know that they are in the control group.  If your treatment comes in pill form, your placebo should not only be a pill, but should be the same size, shape, color, and have the same markings.  If your treatment causes known side effects, like nausea, your placebo should include an ingredient to cause nausea.  What was the "placebo" in the Cheerios studies?

Corn flakes.  Indistinguishable from each other?  Hardly, which brings us to the final criteria.

Were the studies double-blind? 
In a double-blind study, neither the study participants nor the people administering the treatment know which treatment the participants are receiving.  While it is not possible to know the level of blinding achieved without seeing published study protocols, it is inconceivable that the administrators and participants could not distinguish Cheerios from corn flakes.

Whether or not the Cheerios and corn flakes could be distinguished does not necessarily make this a complete fail for these studies.  Depending on the information given to the study participants and administrators, they still may not have known which cereal went to the control group and which went to the treatment group.  Clinical studies regularly assay their quality of the blinding by asking participants  which group they thought that they were in.  In a well blinded study, correct guesses should be randomly distributed.  Without the detailed results of these studies, it is not possible to determine how well blinded these studies were, but combined with the concerns about the quality of corn flakes as a placebo, above, and this quote from a study participant:

After I began eating Cheerios for the study, my cholesterol dropped
by 40
points.

the degree of blinding is an area of concern, especially with the effect sizes observed.

Conclusion
It is highly doubtful that the General Mills'⁹ funded studies meet the rigor required for medical health claim approval by the FDA.  The most likely result is that General Mills will end the specific cholesterol reduction advertising campaign and fall back on the already approved "soluble fiber" claim.

Hopefully, the warning shot fired across the bow of cereal manufacturer signals a bright new day at the FDA.  This Cheerios incident may seem small and trivial, and, relatively speaking it is.  After all, we all know that there is little danger from indulging in Cheerios.  If we are going to say, as we should, that products in the marketplace making claims of health benefits must prove to the FDA that the products live up to those claims, then every product, without exception, must be held to those standards.  And, that means you Cheerios.  As for the FDA, I will be waiting with a frosty, malted beverage ready to celebrate when they also take on bigger fish like homeopathy and Noni juice.

NOTES
1: http://www.fda.gov/foi/warning_letters/s7188c.htm
2: http://findarticles.com/p/articles/mi_m0813/is_8_32/ai_n15691320/
3: According to Google I made this up on my first try.
4: http://healthpsych.psy.vanderbilt.edu/2008/Cheerios.htm
5: http://www.beststuff.com/sportsfitness/cheerios-not-just-for-kids.html
6: RED FLAG - Dr. Gugger formally works for the Bell Institute of Health and Nutrition, which is run by General Mills.  While there is nothing wrong with corporations running research facilities, one should always be skeptical when the research arm is named in a way that does not make its direct connection to the controlling corporation clear.
7: Biomedical research ethics require that, when there is already a standard of care for a condition, the experimental treatment be compared to the current standard of care instead of a neutral placebo.
8: Research into the placebo effect has shown that the expectation of side effects increases the perception of effect from the treatment.
9: Bonus points for any responses identifying the logical fallacy I just used.