Nutritional Pearls: Which Diet Has the Best Evidence?

Matt, a 35-year-old man, is looking to begin following a new diet. However, he isn't sure which would be the best, healthiest option: a low-carbohydrate diet, high-protein diet, the paleolithic diet, a vegetarian diet, a low-fat diet, the Mediterranean diet, or another option.

How would you advise your patient?
 

(Answer and discussion on next page)

Dr. Gourmet is the definitive health and nutrition web resource for both physicians and patients with evidence-based resources including special diets for coumadin users, patients with GERD/acid reflux, celiac disease, type 2 diabetes, low sodium diets (1500 mg/d), and lactose intolerance.

Timothy S. Harlan, MD, is a board-certified internist and professional chef who translates the Mediterranean diet for the American kitchen with familiar, healthy recipes. He is an assistant dean for clinical services, executive director of The Goldring Center for Culinary Medicine, associate professor of medicine at Tulane University in New Orleans, and faculty chair of the Certified Culinary Medicine Specialist program.

It is notoriously difficult to do high-quality nutrition research. The gold standard of research, of course, is the randomized controlled trial, in which a particular treatment is compared to another treatment—ideally a placebo.

As I'm sure you've guessed, there are a few barriers to performing randomized controlled trials with people's diets. First, what if the participants don't follow their assigned diet very well? The ideal solution is to provide them with all of their meals, and that not only gets expensive, it can lead to people dropping out of the study if they don't like the food.

Simply asking people to stick to a diet is fraught with uncertainty, as keeping a food diary to report everything they eat is notoriously unreliable: people like to report that they did what they are supposed to do, so any cheating may be minimized.

This isn't to say that randomized, controlled trials are impossible, but they are comparatively few and far between and tend to be much smaller than the more common prospective studies like the Nurses' Health Study, the Health Professionals Follow-up Study, or the PREDIMED study out of Spain, all of which include thousands upon thousands of people.

A common means of grouping together randomized controlled trials, for even greater quality of evidence, is the meta-analysis. For a meta-analysis the authors identify specific types of research and group the results of these trials together in an effort to magnify the trials' effects, whether positive or negative.

The Research

A group of researchers from multiple regions of Italy came together to perform a meta-analysis of meta-analyses, or as it is described in academic circles, an umbrella review of meta-analyses of randomized, controlled trials.

The authors searched published research articles and identified those meta-analyses of randomized, controlled trials that included adults of at least 18 years of age; identified particular dietary patterns; measured outcomes such as body mass index (BMI); cholesterol scores; glucose, insulin, or HbA1c scores; or blood pressures.

The 80 studies included low-carbohydrate diets, high-protein diets, the paleolithic diet, vegetarian diets, low-fat diets, the Mediterranean diet, the DASH (Dietary Approaches to Stop Hypertension) diet, the Nordic diet, and low-glycemic-index or low-glycemic-load diets.

For each meta-analysis, the authors also analyzed the quality of the included research. The authors state that the included articles made it difficult for them to analyze some research, as the definitions of some diets varied. For example, some of the studied "low-carbohydrate diets" allowed up to 45% of total calories from carbohydrates, while other studies of "low carbohydrate diets" limited carbohydrate intake to 25% of total calories.

Similarly, a "high-protein diet" might be defined by one study as including at least 20% of total calories, while another was defined as including at least 25% of total calories.

"Vegetarian" diets had similar problems, with some "vegetarian" diets being what I would consider lacto-ovo vegetarian and others vegan. It's difficult to evaluate evidence when the evidence isn't actually standardized.

Worse yet, most of the meta-analyses compared their focus diet with "any other dietary intervention, without specific indication" (meaning the comparison could be to literally any other way of eating, from all fast food all the time to bread and water).

Other research into these popular diets had serious flaws: Paleolithic diets, the authors note, were subject to "extensive publication bias, selective outcome reporting, and potential conflict of interests," and for high-protein diets, "the quality of published meta-analyses ... is critically low".

For all the diets included, however, "the Mediterranean diet was the only diet that demonstrated significant and beneficial effects for all the parameters analyzed [from body weight to blood pressures], without evidence of potential adverse effects."

What’s the Take Home?

The current state of nutrition research is not perfect, and this article demonstrates that—although it is improving all the time. Still, this article concludes that the best research we have right now shows that a Mediterranean-style diet is more likely than other popular diet plans to help people improve cholesterol scores, blood pressures, and glycemic scores.

Granted, more and better-quality research is needed on the other diets, but a Mediterranean-style diet is still one of the least restrictive healthy diet patterns.

Reference:

Dinu M, Pagliai G, Angelino D, et al. Effects of popular diets on anthropometric and cardiometabolic parameters: an umbrella review of meta-analyses of randomized controlled trials. Adv Nutr. 2020;11(4):815-833. https://doi.org/10.1093/advances/nmaa006