In this video, a metabolism scientist critiques four recent pieces of research that have been misrepresented in the media to portray low‑carbohydrate and ketogenic diets as harmful or ineffective. He first shows how an epidemiological study labeled a 40 % of calories from carbohydrate diet as “low‑carb,” a level comparable to a Reese’s Peanut Butter Cup, thereby misleading the public about what constitutes a true low‑carb or keto diet. He then dissects a 2025 Science Advances mouse study that used a processed lard‑and‑soybean‑oil chow in genetically obesity‑prone mice, arguing that the diet was not representative of a human keto diet and that the animal model is poorly translatable. Next, he explains how a 2021 Nature Medicine metabolic‑ward trial suffered a fatal design flaw—no wash‑out period—causing a carry‑over effect that falsely blamed the low‑carb arm for metabolic harms observed during the preceding low‑fat phase; a later re‑analysis revealed the opposite effect. Finally, he critiques an N=1 case report in Circulation that blamed a ketogenic diet for rapid coronary plaque progression, despite missing critical details about the patient’s baseline disease and diet adherence. Throughout, he links these examples to broader systemic issues: media’s click‑driven incentives, longstanding anti‑fat bias, and academic pressures that favor sensational but flawed studies. He warns that such misinformation discourages patients and clinicians from considering carbohydrate restriction, which could be beneficial for many. The video concludes with a plug for a newsletter that offers a practical guide to spotting diet‑related misinformation.
Chapters
What if you could live on a diet of Reese's Peanut Butter Cups and in so doing lose body fat and improve your health? Well, if it sounds absurd, it's because it is. It's totally absurd. But that thought experiment reveals a persistent misunderstanding about low-carb and ketogenic diets. Keto. Keto. The keto. The ketogenic diet. Which we're going to get back to, but first I do want to unwrap the real theme of this video. When you think of low-carb or keto, what comes to mind? The sizzle of bacon. The melt of butter. But behind those indulgent images lurks a negative gut reaction. Clogged arteries, strokes, heart attacks. The idea of skinny jeans today being traded in for early gravestone tomorrow. But why? It's not because of evidence, it's because of perception, a feeling. And as a metabolism scientist, I'm fascinated by this intersection between the science and the psyche. So in today's video, we're going to walk through a series of stunning examples to show you just how bad media headlines and The science can bend biological reality to the point that they might even have you convinced that a Reese's cup is low carb. I hope you like horror. They did something tricky. What did they call low carb? It misleads the public and obscures the evidence. This is completely unlike in humans. It's worse than useless. It was that the trial had a fatal error. A low carb diet was blamed for what can fairly be characterized as a lie. Don't let a candy bar So in the rest of this video, we're going to run through 4 examples showing just how badly media and science can bend reality. And believe it or not, each is actually worse than the last. But I promise you, if you can make it to the end, you won't just be entertained— and I hope you like horror— but you will be armed with the sort of depth and rigor of information you need to fight back next time someone throws a nonsense headline your way. So let's get into it. Take as our first example a recent study that was touted on social media as proof that low carbohydrate diets don't help or can even hurt diabetes. At first glance, the headline seemed compelling, if for no other reason than they have the American Diabetes Association seal of approval. But what did the researchers behind this study actually measure? What did they call "low carb"? Well, first, this was a nutritional epidemiological study based on food frequency questionnaires, so participants self-reported their diets. Then researchers divided the participants into quintiles, a fancy word for fifths, according to their carbohydrate intake. But then they did something tricky. They assigned to each group a low carbohydrate diet score. Rather than reporting the percent of calories from carbs or the absolute number of carbs they ate, they gave them this abstract score to make them relative to one another. And here's the problem: the lowest carb group across multiple cohorts was still eating 40% calories from carbohydrates. That's not low carb. It's about the same carb proportion as you find in a Reese's peanut butter cup. Hence the Reese's. Low carbohydrate diets, real low carb diets, are typically less than 20% of calories from carbs. And ketogenic diets, extreme low carb diets, are typically less than 5% calories from carbs, enough to get into nutritional state of ketosis. So to suggest that a 40% 'calorie from carb' diet reflects low carb, it's not just sloppy, it misleads the public and obscures the evidence showing that low carbohydrate diets are effective for both type 2 and type 1 diabetes. And this is not an isolated case, it's just a recent example. So let's take another recent example, another 2025 paper published in Science Advances. This made headlines for claiming that a ketogenic diet causes long-term harm, including fatty liver and insulin resistance. But again, we ask, what is the basis for this alarming conclusion? And once again, it can be summed up in one word: distortion. The study relied on an animal model that was genetically predisposed to develop obesity on high-fat diets, including ketogenic diets. This is completely unlike in humans in terms of metabolic regulation and appetite control. And the truth is even revealed in Figure 1 of the study, if somebody actually reads the paper, which shows that the mice gained body fat compared to controls. And even if we set aside the fact that these were mice, not people, the diet itself was hardly representative of a well-formulated low-carbohydrate or ketogenic diet. It was—brace yourself—a processed lab formulation of lard and soybean oil. I know. So if one were to write a headline that accurately reflected the findings, it might read something more like: In genetically predisposed animals, feeding processed lard and soybean oil chow contributes to fat gain and suboptimal metabolic health. I know, that doesn't exactly grab headlines. Now, before we get to the next study, one that was termed "worse than useless," let's take a moment to step back and understand why misinformation about low-carb and ketogenic diets spreads so easily. You need to look at motivation. Media thrives on clicks and conflict. So "Keto is dangerous," it sells. And add to that decades of baggage of anti-fat bias in our culture, and high-fat diets are just presumed to be guilty until proven innocent. Scientists and reviewers, they're immersed in the same culture, they're not immune. Ultimately that means poorly formulated and designed studies that show harm are more likely to be published, while stronger evidence for carb restriction faces extra scrutiny. The result is a feedback loop. Sensational headlines fuel skepticism, skepticism bolsters academic bias, and the cycle reinforces low-carb diet and keto stigma. And this discourages patients from adopting it, patients who might benefit, and it discourages their clinicians from recommending it. And this brings us to one of the clearest examples of how bias and design flaws combine. An infamous 2021 trial published in Nature Medicine one that was called by Harvard's Professor Walter Willett "worse than useless." It's worse than useless. And on this, I agree with Professor Willett. In this study, what the researchers did is they conducted a randomized controlled feeding trial in a metabolic ward. That's science talk for a very expensive dietary trial with all the bells and whistles. It's the kind of trial that on the surface you want to trust. But what was missed, including by reviewers, was that the trial had a fatal error. Before we get to the study's fatal error, I have my own fatal error, which is that I left the autofocus on on the camera, so you're going to notice a little bit of camera jiggle. I apologize for that, my bad. This is a learning journey. Uh, my editor is going on vacation, so I didn't want to have to make him redo it, so you're just gonna have to deal with it. Stay on your feet. Before we actually get to the fatal error, let's just point out the ketogenic diet in the study was terribly formulated. They termed it an animal-based ketogenic diet, kind of like carnivore, but people were effectively eating fried chicken on salads, broccoli alfredo, all-you-can-eat roasted salted nuts, and whatever the heck that is. Now as any clinician practicing low-carb can tell you, that's the dietary equivalent of an Oreo vegan diet, it's a complete strawman. But whether if the diet design was rubbish by virtue of intent or ignorance is immaterial, because that wasn't even the main problem with the trial. The fatal error in this study was that they didn't include what's known as a washout period, basically a metabolic reset between the two diets, in this case low-fat and a low-carb diet. And this is a huge problem because in dietary studies there are metabolic adjustments that take time when you shift from one diet to another. So if you switch straight from one diet to another, there can be this carryover effect, leading one diet to be blamed for the metabolic harms of the other and/or steal credit for it. And in this case, that's exactly what happened. A low-fat diet stole credit from a low-carb diet and passed blame onto the low-carb diet as well, as was later identified 3 years later published by a separate group. In fact, this carryover effect which wasn't initially reported was about 3 times the original reported effect and in the opposite direction, favoring benefits of adaptation to low-carbohydrate diets. So in short, the findings of the original study weren't just invalid, they were worse than useless because they were misleading. It's worse than useless because it's misleading. They were incredibly misleading. Now unfortunately, the journal Nature Medicine still hasn't retracted this study. I'm not sure why. And the statement in the original trial that still sits there that they looked for such a carryover effect, but the carryover effect was not significant, that was a claim that was entirely unsubstantiated. They provided no tests. So I think, I'm going to make this claim, this constitutes what can fairly be characterized as a lie. But a lie that was endorsed by the peer review process, intentionally or unintentionally, and in this case I think it was unintentional. Now if the true impact of this egregious error, this fatal flaw, isn't apparent yet, let me put it in as simple terms as I possibly can. In this randomized controlled trial, which is seen as a source of truth, a low-carb diet was blamed for the negative metabolic adaptations caused by a low-fat diet and vice versa. And yet, the study still stands as a source of truth— false truth, but truth for many scientists and doctors who don't dig deep enough to recognize the fatal flaw. And you can imagine how the media runs with that. But believe it or not, the examples get even worse. In perhaps the biggest publication lie I've ever seen, a major cardiology journal, Circulation, published a cautionary tale. An N=1 entitled "Rapid progression of coronary artery disease in a phenotypic lean mass hyperresponder who stopped his statin and went on a ketogenic diet." Pause the video here if you need to and reread the title. It makes a very specific claim. Basically, this is a wordy way of saying keto bad, causes heart disease, not good, don't do it. But when you look at this short case study itself, here's the real story. A middle-aged man had severe advanced cardiovascular disease, and at this point in his life he had never attempted a ketogenic diet, to the doctor's awareness. Doctors then went into his heart to unclog a vessel. While there, they saw that another vessel was already partially clogged. So after the procedure, they put him on medications, including statins, to hopefully improve his outcomes. Then he went out into the world for years, during which time his plaque progression was not assessed. And sometime after that, he supposedly tried a "ketogenic diet." But for how long? And what was keto? This information was conspicuously absent. So it's entirely possible that this man already suffering from advanced heart disease developed progression of pre-existing plaque over the years which he was not followed, then he ate bacon for a couple days, calls it keto, and ended with another blocked vessel. Yet the title completely blamed the ketogenic diet. And to make matters worse, the authors actually mischaracterized the patient—he was not even a lean mass hyper-responder. And to reiterate, this wasn't some fringe blog post—it was published in a major cardiovascular journal and presented at a major cardiology conference. So this kind of frank lie becomes truth in academia. I know, I hope you're as appalled as I am. But unfortunately, this pattern repeats and repeats and repeats. These distortions create a steady stream of confusion for the public, and ammunition for critics—ignorant critics—of carbohydrate restriction. The real harm here isn't academic nitpicking, it's practical, because when media headlines distort evidence, patients with diabetes or prediabetes may be dissuaded from trying approaches that could dramatically improve their health. Physicians may dismiss low-carb diets as fads completely ignorantly and inappropriately, and the broader public, overwhelmed by the contradictory reporting, may conclude that nutrition science is hopelessly muddled and give up. This is the tragedy. This is the tragedy of careless reporting, publication bias, and misrepresentation of the data. They prevent helpful messages from reaching the very people who stand to gain the most. Okay, trying to calm myself down a little bit. Um, before we wrap up, I'll give a quick plug. I go into more practical detail in the associated newsletter linked right below. I include a concise practical guide to help you spot future misinformation. But I do now want to do like a Hershey Kiss factory and wrap up. Don't let a candy bar define science. If a Reese's Peanut Butter Cup macros can qualify as low carb on paper, then headlines can say almost anything. And that's the point, that's the point we just proved. We saw it 4 times over. A low-carb score at 40% kilocalories from carbs, mouse chow that was lard and soybean oil sold as keto harm, and a gold-plated metabolic ward trial with no washout that revealed an opposite signal, and a case report that blamed keto for what years of untracked plaque likely did on its own. Now, none of this means low-carb or ketogenic diets are magic. I never said that. It means context and formulation determine outcome. And individuals, not averages, have to live with the results. So next time someone says low-carb doesn't work, ask them to unwrap their claim. Because metabolism doesn't read press releases. Stay curious. I hope this was helpful. And if you were wondering from the intro if I was going to eat this— Heck no. How are there so many of these?
