Health Claims for Food: Good Science, Please

ball_soupaislepic1_5-17_Jeff Keen.jpg

Jeff Keen/flickr

If a shopper who is worried about her risk of heart problems notices a box of cereal that says its whole grains reduce cholesterol, she may be that much more inclined to put it and not another brand of cereal into her grocery cart. Like many people, she knows that cholesterol levels are connected to the chances of having a heart attack.

She may not know that she has just made a biomarker-based decision. Blood levels of cholesterol are one measure used as a biomarker—a benchmark that substitutes for clinical outcomes—for risk of heart attack. How much a customer can trust that cereal's health claim, or for that matter the benefits of a cholesterol-lowering drug, depends on how strongly science connects lowering cholesterol levels with a reduced risk for heart disease.

Use of biomarkers in clinical studies is common and useful, because measuring actual clinical outcomes can be time-consuming and difficult. For example, it's much quicker and easier to measure whether a drug or a calcium-rich diet increases bone density levels than to wait years to see how many people suffer spinal curvatures or broken bones.

But there is a degree of uncertainty in the use of substitutes for actual health outcomes, and regulators and companies must be conscientious about not jumping to conclusions. Every biomarker-based study to support a new claim—whether for a food or a drug—must be thoroughly evaluated on a case-by-case basis. After all, we have been mistaken before. For example, we once thought that blood levels of beta-carotene, a measure of vitamin A, were a good indicator for risk of cancer and heart disease, and that high levels through dietary supplementation were associated with reduced chances for cancer and heart disease. Research eventually showed that some people getting large doses of beta-carotene actually experienced an increased risk of cancer, and none experienced lower risk.

Because consumers naturally presume that health claims for drugs, foods, and supplements are reliable and accurate, these statements need to be well supported by scientific evidence. Statements about the health benefits of food ingredients and dietary supplements have proliferated in recent years. The U.S. Food and Drug Administration has been hampered in its ability to evaluate studies supporting these claims, because it has lacked a method that is broadly accepted across the regulatory, food, and medical communities to rigorously judge just how closely biological benchmarks can predict clinical outcomes. Given that foods and dietary supplements are consumed in such great quantities and can have significant impacts on people's well-being, the biomarker-based studies that food and supplement makers provide to support proclamations of health benefits should be no less rigorously reviewed and evaluated than the studies to support good health outcomes that drug and medical device manufacturers submit.

That's one of the major conclusions of a newly released report from the Institute of Medicine, the health branch of the National Academy of Sciences. I chaired the committee that wrote the report, which provides the FDA—the agency responsible for evaluating medical and health studies—a more comprehensive way to judge companies' biomarker-based studies.

Consumers also need help understanding the science behind the claims they see on package labels. FDA is responsible for ensuring that people are not confused or misled by packaging information. In our report, my colleagues and I call on Congress to make sure FDA has a clear mandate to require necessary studies following initial approval of claims, and that FDA has clear authority to compel companies to revise or remove claims if they are misleading.

Of course, food ingredients and supplements are different from pharmaceutical products and complex medical devices, as is reflected in the different ways these products are regulated. Our committee's recommended method for evaluating biomarker-based studies is designed to work within these different regulatory parameters. The bottom line is that if a company is using a biomarker to generate evidence to support a health claim, there are no rational grounds for using different standards for a food or a medication. Science is science. Consumer's health decisions should be based on good science.