A Tale of “Two Animal-based Keto 🥓 vs. Low-Fat Plant- Based 🌱” Randomized Controlled Trials I want to show you two distinct randomized controlled trials comparing ketogenic diets with low-fat plant-based diets. At a high level, they appear to reach opposite conclusions. 👉One suggests the ketogenic diet has a metabolic advantage. 👉The other suggests the low-fat plant-based diet does. So, let’s compare them and then reconcile the discrepancy. RCT 1⃣: 6 Month The more recent study (PMID: 42660124), which you may have heard of, was a 6-month, randomized controlled trial comparing a ketogenic diet, Mediterranean diet, and low-fat plant-based diet. Participants were provided their meals, and importantly, weight loss was broadly comparable, with each diet producing around 10% weight loss over six months. Despite that similar weight loss, the ketogenic diet produced larger improvements in several metabolic outcomes, including glycemic control, inflammation, hepatic insulin resistance, lipoprotein insulin resistance, and liver fat. The conclusion was fairly straightforward: beyond weight loss itself, a low-carbohydrate ketogenic diet conferred additional metabolic benefits in people with obesity. Or, if you want a quote: “These results demonstrate that a very-low-carbohydrate diet has greater cardiometabolic benefits, particularly in hepatic metabolic function, than matched 10% weight loss induced by a Mediterranean or very-low-fat diet in people with metabolically unhealthy obesity.” TL;DR: The ketogenic diet has an advantage over a low-fat plant-based diet in persons with obesity. Free letter breakdown, here: staycuriousmetabolism.substa… Now consider the other trial. RCT 2⃣: 2 Week Arms, Crossover (with major carryover effect) This was shorter and smaller trial. The animal-based ketogenic and low-fat plant-based phases each lasted only 2 weeks, rather than 6 months. The sample size was also less than half the other trial. Now, there was also another major design difference. Whereas the newer trial used a parallel design, with participants randomized to one dietary arm, the earlier study used a crossover design. That means the same participants flip-flopped from one intervention to the other. In theory, that can be a strength because each person serves as their own control. But it also introduces a potentially major source of bias: carryover effects. If metabolic adaptations induced by one diet persist into the next phase, then the benefits of the first diet can be mistakenly attributed to the second. So, imagine someone follows a low-carb diet for two weeks, develops metabolic adaptations, then switches immediately to the low-fat diet. If those adaptations continue paying dividends during the next two weeks, the benefit created by the low-carb phase could be incorrectly credited to the low-fat phase. And the reverse is also possible: adverse adaptations from the low-fat diet phase could spill over and be blamed on the next. In this trial, that wasn’t just a theoretical concern. That’s EXACTLY what happened. There was a massive carryover effect, large enough that it dwarfed the apparent diet effect itself. (more at this paper: PMID: 38128881) Once you recognize that, the interpretation changes completely. Rather than concluding that the low-fat plant-based diet necessarily produced superior reductions in energy intake and body weight, the data are more consistent consistent with conclusion that positive metabolic adaptations to the low-carbohydrate diet paid off during the low-fat phase, and negative metabolic adaptations from the low-fat phase were blamed on the low-carb diet. Here: pubmed.ncbi.nlm.nih.gov/3812… Now put the two studies side by side. One is a six-month randomized controlled trial with provided meals and the ketogenic diet clearly showed benefits across a range of outcomes, despite weight loss being controlled. The other is a smaller two-week crossover study with a substantial carryover effect that effectively inverts the headlines findings from “people eat less on low-fat plant-based” to adaptations to a low-carb diet provide benefits for reducing energy intake overtime. Once you understand the study design and results beyond the headline, the discrepancy dissolves. 🚨Why this post🚨 But the reason I’m putting these two trials head-to-head is that, superficially, they seem to tell opposite stories. And yet the stronger trial is the one that is seeming to attract more criticism, more caveats, and more attempts to explain away the result that the ketogenic diet produced broader metabolic improvements in people with obesity, even while containing substantially more saturated fat and dietary cholesterol. Meanwhile, the earlier study is frequently cited much more cleanly despite its short during a design problem that should have had it retracted from the literature. That's the problem. If a result aligns with the prevailing narrative, methodological weaknesses can sometimes receive remarkably little scrutiny and get a free pass and unearned defense even after fatal flaws are revealed. When a result challenges that narrative, suddenly every mental gymnastic meaner is deployed to explain it away. I don’t know what to call that other than a double standard.