The diet dog fight: Keto vs Vegan in Nature Medicine

Organisation: Position: Deadline Date: Location:

Although the endless, petty "tribalism and arguments" around diet are tiresome, a recent head-to-head study between low-carb and low-fat diets is worth serious discussion, says Dr Rohin Francis on his YouTube Medlife Crisis channel.

This is a MedicalBrief edited transcript of the broadcast – viewed by close to 200,000 people – by Francis, a British interventional cardiologist, internal medicine doctor, and university researcher, assessing the Nature Medicine study:

A new study published in January pitted a ketogenic, low carb, animal-based diet against a plant-based, low fat diet in a fully controlled environment. A lot of dietary studies are simply not even worth reading, but this one merited sharing, as it was done well. The carbohydrate-insulin obesity hypothesis has been widely accepted but not tested.

Now, I have said before that nutrition research is sullied by a lot of garbage. Untested theories, retrospective questionnaire-based studies that are pointless and inaccurate. That's because it's profoundly impractical to lock people up in a room and monitor everything they eat and do. Well, that's exactly what they did over at the National Institutes for Health in the US. They convinced 21 people to spend a month locked up with all their food provided.

The participants were not told what the trial was about. They were just instructed to do their normal things, not try to lose weight or do anything special. They were actually given loose-fitting scrubs to wear and just eat as much as they want. Half received a keto-style, low-carb, animal-based diet and half received a plant-based, high-carb, low-fat diet, and then after two weeks, they switched.

Now, the first thing that I'm sure you'll say is that this is a month in total. It's too short. Of course, this is a short trial, but this comes down to study design. They're not looking for prevention of stroke or heart attack risk. This wasn't designed to do that. It wasn't powered to do that. The prime outcome for this study was to measure energy and macronutrient intake with a bunch of secondary endpoints like measuring insulin, blood glucose, and things like that.

What was the thinking behind this? The central ethos to low-carb diets, which is the carbohydrate-insulin model of obesity, was the inspiration for this. We know that insulin – and we've known this for decades – has a very important role in how fat is laid down and how it's used up. We know that insulin levels go up when you eat more carbohydrates.

So, carbohydrates cause insulin release and the thinking is that insulin causes fat storage, so the level of circulating fuels in your blood decreases and this tells your body to eat more and to deposit fat and use less energy. This is the idea behind low-carb diets, which remember, have been around for a long time, since the 1970s or so, with Atkins.

Also, it's important to remember that this theory was counterintuitive to the commonly-held theory, which still a lot of people believe now, that it is just purely about total calories in and calories out, and it doesn't really matter what the constituents are. The carbohydrate-insulin model challenged that simplistic belief.

The study in hand randomised participants to either 14 days low-carb or 14 days low-fat and then switched them over. They were all non-diabetic participants. They had an average age of around 30 and an average BMI of about 28, so they were overweight.

The low-carb food was consistent with keto meals. It was animal-based with non-starchy vegetables – things like greens, and the low-fat one was consistent with a plant-based vegan healthy diet, again, with non-starchy vegetables alongside things like sweet potato and so forth.

As expected, the insulin and blood glucose levels were significantly higher in the plant-based group versus the keto group and blood ketones were higher in the keto group. So far nothing new, nothing that surprising. So why am I telling you about this trial?

Because people ate significantly less on the plant-based diet right from the beginning, and even with the documented high insulin levels – which we would think would tell them to eat more, that's the belief – they ended up eating less, and consumed about 700 calories less per day. Quite strikingly, this was seen in every single participant. There was no noise – it was every single participant from Day 1. This is simply because plant-based diets and plant-based foods are just lower in calories per gram, so you feel full and you've eaten less calories.

There were no differences in hunger, fullness, satisfaction, and eating capacity. Interestingly, things like familiarity and enjoyment were also the same as well, which is part of the selection criteria. They didn't choose fussy eaters and they chose people who would be familiar with most of the food that they were presented with.

With the lower caloric intake of the plant-based diet, people lost more body fat, not on the keto one. They lost more water weight on the keto diet, which is a recognised phenomenon that occurs early on, and this is to do with glycolysis activation mobilising water, but body fat was not significantly reduced with the keto diet. Overall, in spite of the keto group eating 700 more calories per day, they lost more weight overall, but I reiterate this was not body-fat weight loss.

So, should we chuck out the carbohydrate-insulin model like a tub of E-number-enriched margarine? Well, a lot of people think we should. Are the effects on insulin and glucose in the keto diet and the lower levels of particular interest and importance to metabolic health and diabetics? Probably yes. Is a healthy plant-based diet a good starting place for good weight loss? I think it probably is.

As I said at the start, the most tiresome thing to people who aren't staunchly in one of the different camps is the endless arguments that assume that life is binary and you can only do one diet or the other and you can't possibly be flexible in your choices, but can't we all just get along, because there is actually some common ground here?

If you look at the food that they served in this trial – there are some pictures on the NIH website – both groups had meals that look really nice, lots of non-starchy veg and a distinct lack of ultra-processed food. Keto and plant-based diets might seem worlds apart, but are they really? They both emphasise moving away from the standard US diet, which we all know is awful.

That's the overwhelming message that emerges from a multitude of trials, including work by this same group, that we can argue till the vegan cows come home about all the different macros and ketones and so on. But the ultra-processed foods like ready meals, and cookies, and crisps, and frankly, most things I see in US supermarkets. The UK is not far behind. We're both significantly worse than in mainland Europe. But foods with long lists of additives, refined carbs, salt, and so on, they're the real problem.

I read somewhere – don't quote me on this – but something like two-thirds of the US diet comes from ultra-processed food compared to about 5% in Portugal and 8% in France. I think UK was somewhere in the middle, around 30%.

This trial is an important extra data point. By all means, criticise the duration, but people complaining about the sample size are missing the point here. Both groups lost weight.

Instead of fighting between belligerent camps, or more importantly, instead of berating yourself for not sticking to a certain diet and messing up your ketosis, or being unable to stick to a diet that your friend promises you is the best one, just try one intervention first. That's cutting your intake from highly-processed foods down to as close to zero as you can.

I appreciate that the term "highly-processed" is also confusing and that needs a bit of clarification as well. But whenever people talk about that healthiest populations in the world in Italy or Japan, none of them are doing extreme, restrictive diet, and there's little in common between a lot of these places in terms of what they're actually putting in their food, except there's one thing that is in common. They are mostly natural, good, unprocessed foods.


Medlife Crisis YouTube: Diet Dog Fight: Keto vs Vegan


Low-fat, plant-based diet to low-carb, animal-based diet compared — NIH study


Nature Medicine study (Restricted access)


See also MedicalBrief archives:

Low and very-low carb diets for type 2 diabetes remission – Systematic review


‘Compelling’ evidence for a ‘good-fibre’ diet over a low-carb diet


Review assesses low-fat/high-carb and high-fat/low-carb diets


Both low-carb and high-carb diets raise risk of early death — 25-year study


Endocrine Society: Both low-carb and low-fat diets work


Receive Medical Brief's free weekly e-newsletter

Related Posts

Thank you for subscribing to MedicalBrief

MedicalBrief is Africa’s premier medical news and research weekly newsletter. MedicalBrief is published every Thursday and delivered free of charge by email to over 33 000 health professionals.

Please consider completing the form below. The information you supply is optional and will only be used to compile a demographic profile of our subscribers. Your personal details will never be shared with a third party.

Thank you for taking the time to complete the form.