Anyone who follows a keto lifestyle is no stranger to hearing how terrible it is from uninformed and unqualified pundits on social media. (Jillian Michaels, anyone?)
A recent editorial in JAMA Internal Medicine, however, is written by experienced, respected physicians, and it calls into question the evidence supporting the "hype" of keto diets. These authors certainly are not unqualified, and they are not uninformed, at least not in other areas of nutrition science.
JAMA Internal Medicine: The ketogenic diet for obesity and diabetes - enthusiasm outpaces evidenceThe authors do, however, reveal some clear biases. While I applaud their efforts to focus on evidence, I am afraid they missed the mark.
I will address many of their points in this post, but for a deeper discussion of the science backing keto, I encourage you to read our guide "The science of low carb and keto." It contains dozens of links to studies supporting the use of low-carb and keto diets for weight loss, diabetes, liver disease, PCOS and more.
Getting back to the JAMA editorial, a crucial point that we cannot ignore is that the authors are known vegan advocates and have commercial interests in promoting a vegan diet; all three are associated with the vegan promotional film "Forks Over Knives." While that does not invalidate the evidence they cite, it may explain why - out of the vast literature on low-carb diets - they chose the evidence they did. Their dietary biases certainly cloud their judgement and unfortunately keep them from seeing the broader picture. This is clear right away from their language.
For instance, they refer to the keto diet as requiring followers to "forgo nearly all carbohydrates." I would have phrased this as allowing followers to "eat almost endless amounts of above-ground veggies as their source of real-food, natural carbohydrates."
It turns out the words we use matter, as do our conflicts of interest.
But what about the science? For starters, they fall back on the mantra that all we need to do to lose weight is restrict calories. How can we blame obesity and metabolic disease on a low-fat diet when statistics show we eat more calories than we used to?
This type of thinking completely ignores the human aspect of food and weight loss. Following a diet that we enjoy, that keeps us full, and reduces our cravings is the key to reducing caloric intake. We aren't calorimeters. We are people with emotions, cravings and hunger.
It also ignores the fact that both the timing of our eating and the degree of insulin stimulation from carbohydrates contribute to weight loss (or lack thereof). Although overall calories do matter, they do not exist in a vacuum. The quality of those calories matters just as much, if not more than quantity.
Next, the authors refer to a 2013 meta analysis of randomized controlled trials (regarded as the highest level of evidence) that shows low-carb diets lead to greater weight loss than low-fat diets. However, they question the significance of the 1 kilogram weight difference. That is fair, as 1 kilogram is not much. But they neglect to mention that the low-carb group also had significantly better blood pressure, triglycerides and HDL (also shown in this 2016 meta-analysis). That would certainly suggest that either the weight loss difference was clinically significant or that there were additional benefits to low-carb diet beyond weight loss. Those are important conclusions from the study that an unbiased assessment would not have overlooked!
This next part surprised me. The authors suggested that there is no quality evidence to support low-carb or keto diets for diabetes. I guess the American Diabetes Association disagrees as it recently amended its recommendations to include low-carb and keto diets as an evidence-based treatment option to control blood glucose and diabetes.
The authors also called into question the Virta Health data showing improvement and even remission of type 2 diabetes with keto diets out to two years. They are correct when they pointed out that this trial was not randomized and included interacting with a health coach via a smart phone app.
Regardless, we cannot ignore the fact that the Virta trial showed better remission rates and medication elimination rates than any dietary study in the medical literature. While the keto group may have received more support, based on this study, there is no debating that keto eating was part of the most impressive diabetes reversal program in the medical literature. For the sake of millions of patients struggling to manage type 2 diabetes, this point needs emphasis.
I am particularly interested in the statement claiming it is more important to control glycemic levels while consuming "healthful carbohydrate rich foods...even in the absence of weight loss." In an editorial promoting the importance of evidence, this statement curiously lacks a reference and sounds quite far-fetched to me.
There are plenty of other issues as well. The authors pointed to keto diets in kids that have led to nutrient deficiencies and rare complications. Most of these diets were made up of shakes and meal replacements rather than food, and therefore do not apply to modern-day, real-food keto diets.
In addition, they mention a laundry list of "complications" that are either short term (muscle cramps), mild (constipation), or simply not true (bone fractures and "multiple vitamin and mineral deficiencies").
See our guide to low-carb and keto controversies here.
But perhaps the most glaring concern with this editorial is the statement that keto diets come at an "opportunity cost" of avoiding "healthy" whole grains, fruits and legumes. Once again, we need to look at the data more carefully and not get enticed by the headlines.
The majority of the trials showing benefit to whole grains compared them to refined grains. It shouldn't be a surprise that whole grains are better than the refined grains found in processed food. Yet in the editorial, the authors imply that eating whole grains is healthier than eating a real-food, low-carb diet without grains. That is not an evidence-based claim, as that study has never been done.
The other set of data in support of whole grains, fruits and legumes comes from uncontrolled observational studies. For insulin-sensitive individuals with other healthy habits, such as those in "Blue Zones," these foods can be part of an overall healthy lifestyle. It doesn't mean they are the main cause of good health, it simply says they can be part of that lifestyle.
All we need to see is the glucose spikes from a patient with diabetes who eats "healthy" whole grains and fruits to know that these foods affect different people in different ways. Opportunity cost? I am more concerned with the opportunity cost of people missing the powerful tool of low carb because they think they need fruits and whole grains to be healthy. That is simply not true.
(See our full guide on whole grains here.)
In the end, while this is a thoughtful rebuttal against a keto diet, it still presents evidence from an overly biased perspective and fails to see the whole picture. Low-carb and keto diets are absolutely supported by the literature to help with weight loss and the reversal of type 2 diabetes.
That's my bias. I strive every day to acknowledge it and present information as objectively as I can. That's why I encourage everyone to read our science of low carb and keto guide with extensive references to support the health benefits of low-carb living. Low-carb and keto diets are a powerful tool for health, and we have to see past the biases that some experts bring to a review of the literature in order to help our patients and help ourselves.
Thanks for reading,
Bret Scher MD FACC