Amazon lists 60,000 books under the search term diet. That many books on the same subject tells me that there is no accepted standard. Different books propose different styles of eating: “Avoid carbohydrates!” “Eat intermittently!” “Eat like a caveman!” “Increase foods that are antioxidants!” The list of recommendations is vast.

Assuming that “healthy eating” refers to a diet that will allow you to avoid disease and enhance your strength and vigor, your search is in vain. Why? Because accepting existing reports of healthy eating on a similar basis to other reports on health-promoting, or unhealthy behaviors, misses an important difference. To prove the effects of any one of the numerous published “healthy eating” proposals requires information that is not available.

Let’s review how something is proved. In medicine, the gold standard is the controlled clinical trial. Here, two or more groups are stratified by some blind, randomized process into what seem like equally eligible participants. Each group gets a unique treatment: one may get one drug while one may get another or may get a placebo. When the populations are compared after a predetermined period of time, any difference between them must have been due to their specific treatment. There are no random events to confuse the conclusions.

To carry out such a trial for a dietary plan would require two components: At least two groups were randomly chosen, each of which followed a different food regimen. While this sounds simple, it is not feasible for very long. Only by using populations without free will, such as the military or prisoners, could a test diet be imposed. That brings up bioethical concerns. Free-living people are not likely to stay exactly on their mandated food program (Lichtenstein et al. 2021). For how long would the study have to be carried out? To answer that, one must know how long it takes for something unhealthy to show up; that is usually a long time, unless one diet contains something acutely toxic. Keep in mind that cigarette smoking—perhaps the single most serious deterrent to good health—takes ten to twenty years to show harm. So, consider “eating healthy” food plans. None have any ever been tested that long.

The most commonly discussed “healthy diet” is the “Mediterranean diet,” which represents generations of people but contains no randomization. The United States Department of Agriculture used this diet concept in their eighth edition (2015–2020) of dietary recommendations (U.S. Department of Health and Human Services and U.S. Department of Agriculture 2015). The problem is that people who live around the Mediterranean are different in many ways from people who live elsewhere. Their genetics, lifestyle, health care, and too many other variables to list here are almost certainly different from populations that live elsewhere. How can one decide that from among all these differences that their diet is the most important aspect of their health status? Beyond that, their diets have changed enormously over the decades, so historical results may not be sustained. The Mediterranean diet is from an epidemiological study, not a controlled randomized trial. As such, it cannot be accepted as factual without many caveats.

Epidemiology is the branch of medicine that deals with the incidence, distribution, and possible control of diseases and other factors relating to health. A recent example of how useful epidemiology can be occurred during the COVID-19 pandemic (Anderez et al. 2020). From information such as infection rate, contact frequency, and prevalence of coexisting conditions, the spread of the disease was accurately predicted. No information related to the Mediterranean diet exists. All epidemiological studies include questions such as: Have the results shown a cause-effect relationship?

Showing that an exposure is strongly associated with a disease does not necessarily imply that there is a cause-effect relationship. What is the strength of association? Is there a temporal relationship between exposure and outcome? Is there a dose-response relationship, consistency, and biological plausibility (Zaccai 2004)?

While probably hundreds of properly randomized clinical trials (RCTs) of food regimens have been carried out, their limitation is they last only from weeks to months—much shorter than the time needed for most diseases to appear. Such studies examine targeted parameters of health, such as markers of inflammation, lipid levels, and blood sugar, but diseases are not seen in this short time.

Lichtenstein (2021) comments:

In some human nutrition RCTs, “hard endpoints” or disease outcomes, such as incident events (e.g., stroke, myocardial infarction) or disease status (e.g., carotid artery thickness, arterial calcium score) are prespecified. Human nutrition RCTs are less likely to use hard clinical outcomes than in other types of studies because, in addition to the long lead times for natural disease progression, there are other challenges, such as sustaining adherence to a dietary modification (avoiding recidivism). Hence, long-term human nutrition studies are most often observational in nature.

Equally important, it is rare for trial reports to follow up on whether any weight lost in the trial was maintained when the participants left the trial or what their future health was like. It is generally noted that few people maintain trial-related effects when they are again eating without restraints.

Indeed, there are some known diet-health relationships. For diabetes, control of blood sugar through restrictions in carbohydrates is critical. For some hypertensives, salt restriction is important. But, for most diseases, diet and the condition have no short-term relationship.

I posit that the diet-health linkage we see today is a reflection of the diet-health linkage from long ago. It is critical to recognize that diseases such as rickets, scurvy, goiter, and certain anemias are deficiency diseases that can be prevented by dietary improvements. Centuries ago, our food supply was not as abundant or complete as it has become starting in the twentieth century. Perhaps our recollections of these deficiency diseases keep the vitamin shelves in our pharmacies loaded with a dizzying variety of products. However, unless one is eating a diet with insufficient variety to provide all the vitamins and minerals our body requires, multivitamin supplements are not needed (NHS n.d.). But this historical mindset may be why so many of us believe in the idea of “healthy eating.” Ergo, the 60,000 diet books.

To clarify, I want to differentiate the terms diet and food regimen. We associate dieting with weight loss. Nutrition is much more concerned with our health than with our weight. A food regimen is a set of meals not necessarily designed to help us lose weight but to maintain our health by giving us all the nutrients in the right proportions. That should be everyone’s goal.

It is very likely that “healthy eating” is no more complex than keeping a healthy weight. But we know that is not easy! If you want to “eat healthy,” have a balanced diet, and don’t fill your plate so full. It is less what you eat but how much!

References

Anderez, D.A., E. Kanjo, G. Pogrebna, et al. 2020. A COVID-19-based modified epidemiological model and technological approaches to help vulnerable individuals emerge from the lockdown in the UK. Sensors 20(17): 4967.

Lichtenstein, A.L., K. Petersen, K. Barger, et al. 2021. Perspective: Design and conduct of human nutrition randomized controlled trials. Advances in Nutrition 12: 4–20.

NHS. N.d. Do I need vitamin supplements? Online at www.nhs.uk.

U.S. Department of Health and Human Services and U.S. Department of Agriculture. 2015. 2015–2020 Dietary Guidelines for Americans, 8th Edition (December). Online at http://health.gov/dietaryguidelines/2015/guidelines/.

Zaccai, J.H. 2004. How to assess epidemiological studies. Postgraduate Medical Journal 80(941): 140–147.