The observation that mice fed diets high in polyunsaturated fatty acids (PUFA) versus saturated fats become obese, while humans are often advised to consume PUFA for health, certainly seems paradoxical at first glance. Let's break it down with a skeptical lens.
1. **Species Differences**: *Hmm, mice aren't tiny humans. Their physiology is different.* Mice and humans metabolize fats differently. What causes obesity in mice might not have the same effect in humans. For example, mice have a higher metabolic rate and different lipid processing mechanisms.
2. **Types of PUFA**: Not all PUFA are created equal. Omega-3 and omega-6 fatty acids are both PUFAs, but they have different effects on health. Omega-3s are generally considered anti-inflammatory and beneficial, while excessive omega-6s (which are more common in Western diets) can be pro-inflammatory.
3. **The Role of Diet Composition**: *It's not just about fat types, but the whole diet.* In studies, the context of the entire diet matters. Mice studies often use highly controlled diets, which don’t fully replicate the complexity of human diets. The impact of PUFA might differ in the context of a balanced diet containing fiber, protein, and other nutrients.
4. **Quantity and Quality**: The quantity of PUFA and the overall diet quality in humans matter. Excessive calorie intake, regardless of fat type, can lead to obesity. Also, processed foods high in PUFA may have other unhealthy components.
5. **Long-term vs Short-term Effects**: Mice studies often look at short-term effects, while human dietary guidelines consider long-term health outcomes. Short-term weight gain might not correlate with long-term disease risk.
6. **Generalizing from Animal Studies**: *Mice studies give clues, but they're not conclusive for humans.* Animal studies are a starting point for understanding biological mechanisms, but their findings need to be interpreted cautiously when applied to humans.
7. **Epidemiological Evidence in Humans**: Observational studies in humans have shown that diets high in certain types of PUFA (like omega-3s) are associated with reduced risk of heart diseases and other health benefits.
8. **Clinical Trials**: Controlled trials in humans provide more direct evidence. Many have shown benefits of PUFA-rich diets in reducing risk factors for heart disease.
In summary, while mouse studies are valuable, they can't be directly extrapolated to humans due to physiological differences. The current dietary recommendations for PUFA intake in humans are based on a large body of research, including epidemiological studies and clinical trials, which generally indicate health benefits, especially when replacing saturated fats with PUFA. *It’s a complex picture, and dietary advice must be nuanced.*
Hi Borge, please keep commenting here, you sound like a man who can show me when I'm missing important things, and that's my favourite kind of comment.
I very much agree that 'PUFAs are bad for mice' doesn't straightforwardly imply that 'PUFAs are bad for humans'.
In fact for nearly a year now, I *thought* that I was arguing that 'PUFAs are not bad for mice' didn't imply that 'PUFAs are not bad for humans'.
Which I was perfectly happy to do, mainly because I figured mice would be more tolerant of PUFAs than humans because of their natural diet.
But I *thought* that I had to explain why the PUFA thing might be true *in spite* of mouse studies.
Whereas in fact it seems that the mouse studies might be on my side.
And at that point, the whole 'seed oil hypothesis' just looks like a racing certainty to me. There's no mystery to explain, it can all just be straightforwardly true, and there's no real evidence against.
I'm shocked and surprised by this, and I'm still having trouble believing it. I keep thinking some pro-seed-oil person is going to leave me a comment with twenty studies proving that mice do fine on seed oils.
I would have thought that before people went around telling everyone about the benefits of polyunsaturated fats for seventy years or so, they would have at least checked to see whether they poisoned mice. That's the first thing we tend to do when testing to see whether things are safe to eat isn't it?
I wasn't expecting them to have done it well, but I was expecting them to have had a pop. It just never occurred to me that they hadn't bothered.
The progression from observing effects in animals to making dietary recommendations for humans isn't direct or simplistic. In the case of PUFAs, the shift towards incorporating them in human diets wasn't solely based on animal studies. It also involved epidemiological evidence in humans, observational studies, and controlled trials. Did researchers jump the gun? Perhaps not. These human studies generally showed that diets rich in certain types of PUFAs (especially omega-3 fatty acids) were associated with lower risks of heart disease and other health benefits.
Why Still Use Rodent Studies?
Rodent studies remain valuable for several reasons:
Basic Biological Insights: They offer insights into basic biological processes and mechanisms that might be relevant to humans.
Controlled Environment: Rodents can be studied in a controlled environment, allowing for the isolation of specific variables.
Ethical and Practical Reasons: It’s often not feasible or ethical to conduct certain types of experimental studies in humans.
Applicability of Rodent Studies to Humans....This is the tricky part. Not all mechanisms observed in rodents apply to humans.
The decision to consider a mechanism applicable depends on:
Biological Plausibility: Does the mechanism make sense in the context of human biology?
Consistency with Other Evidence: Does the observation align with epidemiological and clinical evidence in humans?
Further Research: Are there follow-up studies in humans that support or refute the findings from rodent studies?
The recommendation to include PUFAs in human diets wasn't primarily for obesity prevention but rather for overall cardiovascular health. While obesity is a risk factor for many diseases, the type of fat in the diet also plays a role in cardiovascular health, independent of obesity. It’s a balancing act. Recommendations for PUFA intake are part of a broader dietary context, emphasizing moderation, variety, and the replacement of less healthy fats.
Critique and Evolution of Dietary Guidelines: It's important to note that dietary guidelines evolve as new research emerges. Early recommendations may have been based on the best available evidence at the time, which has since been refined. Science is always a work in progress.
Indeed, not all PUFAs are the same. The two main types are omega-3 and omega-6 fatty acids. While both are essential fats that the body cannot produce on its own, they have different roles and effects:
Omega-3 Fatty Acids: Found in fatty fish, flaxseeds, chia seeds, and walnuts, these are generally considered anti-inflammatory and beneficial for heart health.
Omega-6 Fatty Acids: Common in many vegetable oils, nuts, and seeds, these are also essential but can be pro-inflammatory when consumed in excess, especially relative to omega-3s.
The context in which PUFAs are consumed is crucial. In ultra-processed foods, PUFAs are often present in refined oils and may be accompanied by high levels of sugar, salt, and unhealthy additives. The negative health impacts of ultra-processed foods could outweigh the benefits of PUFAs.
There's a growing understanding that the matrix in which nutrients are consumed matters. For instance, nuts and seeds contain fiber, vitamins, and other compounds that may synergistically contribute to their health benefits. In contrast, extracted oils are more concentrated in fat and lack these additional nutrients.
Comparing the fatty acids in cheese, milk, and butter is an important point – the overall dietary matrix. Recent research suggests that the health effects of dairy fats may depend on the other components of the dairy product and the overall dietary pattern where butter often elevates LDL and is considered risky when consumed in excess, whereas cheese, fermented dairy and milk is considered healthy/net positive.
While certain PUFAs are beneficial, they should be consumed as part of a varied diet rich in whole foods. Overconsumption of any type of fat, even healthy ones, can lead to negative health outcomes.
I think many are taking it too far when they obsess about low-PUFA pork or poultry if otherwise eating lean meats, and avoiding all nuts and seeds - when ALL relevant research considers these to be beneficial for a range of health effects. The average american consumed too many calories overall, where PUFAs are contained in the ultra processed junkfood they are eating. Not eating walnuts *because of PUFA* is missing the forest for the trees.
Borge, thanks for everything you say here, it's all perfectly sound and true.
But do you seriously think it was wise to recommend putting PUFAs in food without checking to see whether they caused any problems in mice?
How would you react if some quantum holistic shyster recommended putting some aromatic compound derived from magic healing crystals in the food supply in very large quantities without even doing the most basic safety tests?
I honestly don't know the timeline of this, but again - mice aren’t small humans, many things that was bad for mice was neutral in humans, many things that were amazing in rodents were bad in humans etc etc.
Check out the Norwegian studies in the Tucker Goodrich link (https://tuckergoodrich.substack.com/p/does-linoleic-acid-induce-obesity). Maybe I'm wearing the omega-6 tinted glasses here, but it seems like pretty much exactly the study you're asking for - mice on 1% vs 8% LA, both on a high-fat diet (same total fat amount).
Maybe I'm being full of shit here, and reading something into it that isn't there. If so, please let me know.
I think the #1 reason that it's not an instant open/shut case is that it is indeed somewhat complicated:
1. There's a buffer/reservoir thing going on, so just stopping seed oils now doesn't guarantee you'll lose weight tomorrow
2. There's almost certainly a genetic or similar component, where it affects people to a different degree, though almost nobody seems to be entirely immune (<8% of U.S. population has 0 metabolic syndrome symptoms)
3. There is at least the interaction with BCAAs/isoleucine, maybe others. Again this doesn't seem to affect all people. It'd be interesting to see the individual data for the mice in those isoleucine studies - did they all respond the same, or did we have some lean mice and some super obese American mice? In fact, I think I'll ask Lamming if he has that data.
> Maybe I'm being full of shit here, and reading something into it that isn't there. If so, please let me know.
No, I'm beginning to think you're just straightforwardly right. But we need an intelligent enemy to tell us what we're missing. I wonder where we can find a clever pro-seed-oils man who'll give us the time of day?
Could try Stephan Guyenet? Maybe he'll answer your emails. He used to be anti seed oil (he turned Tucker Goodrich!) but then suddenly changed his mind for unknown reasons.
Yes, I've heard of him. If he used to think what I currently think and now doesn't I need to know what changed his mind. I wonder if he wrote it down anywhere?
Did you think this whole time that we were in a "race to find the truth" with Big Science? Of course not. Big Science has no interest in finding out what causes obesity; it'd ruin the whole racket.
It is trivial to test this. It has been tested. A child could do it.
And yet, it's been swept under the rug. When I talk to Actual Serious Scientists about it, they pretend they didn't understand the question or misheard me or radio silence.
I talked to Lamming, the guy who runs the lab with all those isoleucine studies, about this - hey, all your mice are on a high-PUFA diet, even the control mice have 3% LA and seem prediabetic from the glucose tolerance test!
He said he thinks his mice are only 1-2% LA. Turns out he didn't know how much LA is in corn oil, the main fat source in his control chow.
They wear CICO/fat tinted glasses. They cannot see the truth if you hit them over the head with it.
After over a decade on keto, following a bunch of interactions between ketoers and Big Science, I believe very strongly that a lot of mainstream scientists are straight up crooks. They're dishonest, they will lie, they will misrepresent their own findings. Maybe they believe their own lies, though it's hard to see how that could be possible.
Kevin Hall, THE big shot in U.S. nutrition science, likes to post a study of his titled something like "All diets work exactly the same," the title image (!) of which shows that low-carb diets work 30-70% better than low-fat diets.
He also published a study saying "high-fat ketogenic diet is obesogenic." The study: he compared a junk-food "high-fat" (=mixed) diet to a low-fat, practically vegan diet ("Whole Food Plant Based"). The high-fat people ate like 700kcal/day more (IIRC) yet lost the same amount of weight!
Physics defied! First law of Thermodynamics broken! CICO disproven! Right?!
No. Obviously, his finding was "Ketogenic diet is obesogenic by causing increased caloric intake."
He is simply a liar. Maybe he believes his own lies, which would make him a zealot, but there's no functional difference.
You seem to believe in the Efficient Market Hypothesis of Science - if it was easy to demonstrate, it would be common knowledge. I believe the Market for Science is completely rigged, and it is not at all efficient.
There are tons of these experiments. Grant from https://ggenereux.blog/ literally did the experiment you mentioned, just with vitamin A - he bought rodents, put them in cages, and didn't feed them any vitamin A. They never, ever got vitamin A deficiency. It's so trivial a child could do it, yet science "knows" you couldn't possibly survive for more than a few weeks without vitamin A.
I do actually think that if we find something that is reliable and easy to demonstrate, it won't just get ignored!
We have to actually do that though. At the moment, we've got you and me claiming to have lost considerable weight, and me claiming that my whack-job thyroid stuff is getting better, but neither of us are claiming to have got back to normal health and weight. In fact both of us are having real trouble. And that's about it.
I think that it all makes a certain theoretical sense, but even you don't take me seriously about all that. And you're right not to. Unless my theories end up showing results and predicting things, which at the moment they're not doing.
That's enough to get me interested, but it's not nearly enough that I could convince anyone sceptical that we'd found something important, or even to try it. Any fad diet has that kind of evidence, and they're mostly bollocks.
But if it becomes viral and popular, as it inevitably will if it actually works and we can convince people who are already open to trying things, then eventually a mountain of anecdotes will become real evidence, and then people will start to try to test it formally, and eventually it will win.
And one day, what is actually going on, right down to the molecular biology of it will be settled science.
Our job is to get it to work on us, first. And then if we can get it to work on us, to get it to work on other people.
I almost totally agree for once, except that it doesn't actually need to work to get to stage 3.
If we're right, then (no-PUFA) keto and carnivore will actually work, maybe real slowly, but the fact that they've got to stage 3 doesn't convince me that they work. It just sort of hints at it. And science is right to be very sceptical of them. You yourself got very fat on a keto diet!
I think it kinda, sorta needs to work to a degree.
Low-carb took off and it works in maybe 20-30% of people (worked better before the PUFA content increased so drastically, I imagine).
So did keto, which works in 30-50% of people, carnivore maybe 50-70%.
So we don't need to get to 100%, but it needs to work at least a little bit.
I did gain 100lbs on keto, but I also lost 100lbs on keto. It DOES work, just kind of orthogonally if you accidentally do it right ;) Carnivore has a higher hit ratio, because the only PUFA you can eat on it are chicken and pork, and it won't work if you're one of those low-protein people like me, but that seems to be maybe 30-50% of those who try it only.
Kate was telling me about her teenage experiences with the F-plan diet. Very popular at one point. Jesus Christ. Young women really will do just about anything to themselves to lose weight.....
> You seem to believe in the Efficient Market Hypothesis of Science - if it was easy to demonstrate, it would be common knowledge. I believe the Market for Science is completely rigged, and it is not at all efficient.
Depends on the science.
In fundamental physics, I do believe EMHS pretty much. If you can find something that routinely seems to violate the currently understood laws of physics (which we know are wrong!), then physicists will be very very interested, and they will try to pin down the precise mechanisms, and if necessary, they will upend their whole world view to account for the tiny detail.
In English Literature, say, there are no laws, anything goes, nobody cares. They don't make any claims that could be tested, they don't even believe in truth. I see no reason at all to trust them about anything, although I can sometimes see a certain value to some of their work.
The question is:
Where on the spectrum of all the things that get done in universities does medical stuff lie?
And I think that things like molecular biology are trying very hard to be the first type of thing, and having great success.
And I think that things like nutrition advice and psychology seem more like the second type of thing.
They're getting better though. People take the 'replication crisis' very seriously, and are trying to reform the system so it more reliably leads to finding out true things. But it's really hard. Managing large scale collections of people is always really hard. Changing systems is really hard.
Yea it's definitely closer to the English Lit Major end of the scale.
I actually think it's gotten much, much worse. Our state of knowledge in nutrition is significantly worse than it was 100 years ago - we now un-know several true things, and we know a lot of falsehoods.
By the way, last night I was lying awake thinking about mice (as one does) and it dawned on me: of course there's a somewhat mainstream (in carnivore/keto circles) explanation for why even the low-PUFA, "high-fat" mixed diet causes obesity: The Randle Cycle. The Swamp.
These diets are typically 45% fat, and therefore still high in carbs. The "chow" diets are often 10% fat, rest mostly carbs. So a fair comparison would be a ketogenic 10% carb diet.
This still doesn't explain why the Norwegians managed to keep their mice lean on a 1% LA, 60% fat diet.. is 60% fat already enough in absence of LA? Or is there something else we're missing?
> Did you think this whole time that we were in a "race to find the truth" with Big Science?
Yes, that is what I thought. The only reason I thought we had a chance is that they are so useless.
It is what I still think. If we've come up with a simple cheap technique that works to fix obesity and all the other diseases of modernity (which I *can* believe, but do not yet believe), then it will spread like wildfire, and there is nothing anyone is going to be able to do to stop it.
If it all turns out to work then you, I think, were the one that made the crucial observation. No way on earth will you ever get the credit for it, but we will know. And I can feel good about having been able to spot the one man talking sense in all the gibberish.
Do you think it's a coincidence that tens of thousands of professional scientists funded by taxpayer money apparently haven't figured out what a pig farmer from upstate New York and some asshole like me and a boat-dwelling, choir-singing, chips-frying bon vivant from Cambridge managed to figure out in a handful of years?
They aren't even looking for it. Of course there are a tiny % that are doing helpful stuff and go unnoticed, like Lamming. But it's rotten at the top, much of the way down, I think.
edit: And yes, if it works good enough (70%?), it might spread and take over and solve the issue (maybe to 70% or something). The reason it hasn't is because this is only true as of social media.
When I first heard of keto, which was already 20 years or so into the internet, it was that dangerous evil diet nobody could possibly survive, and which would almost certainly kill you. Carnivore wasn't even a thing people could fathom.
> Do you think it's a coincidence that tens of thousands of professional scientists funded by taxpayer money apparently haven't figured out what a pig farmer from upstate New York and some asshole like me and a boat-dwelling, choir-singing, chips-frying bon vivant from Cambridge managed to figure out in a handful of years?
Well how many of them were actually thinking about the problem?
The existence of standard breeds of lab mice and standard mouse diets shows that they're trying to be careful.
I'm interested in "PUFAs bad" on the basis of evolutionary thinking that makes sense to me, but that no one else takes seriously at all, and in ex150 originally because your original arguments made intuitive sense to me, and now also because it does seem to work well for me.
Brad is interested in "PUFAs bad" because of molecular biology that I don't understand nearly well enough to evaluate, and on the basis of torpor/hibernation theories that I think are risible. And why he's suddenly got interested in BCAAs is beyond me, but it rhymes quite nicely with the protein restriction bit of ex150.
If we can raise "PUFAs bad, protein involved" to the level of a hypothesis that people take seriously at all, then suddenly all those diets will get their various lipid levels measured and controlled very accurately, and then they'll have the tools to actually discover stuff.
Sorry, I mean, "Thinking about the problem through the lens of PUFAs bad". There hasn't been anything to raise it to their attention out of the utter swamp of plausible competing ideas.
I've suspected all my life that saturated fats and salt are fine, on exactly the evolutionary grounds that you think are dangerously misleading.
I've wondered whether PUFAs are actually actively harmful for about the last year. There was nothing to raise it to my attention either, and I'm unusually interested in crackpottery in all its forms.
It seems to me that that's it? But trawling through the medical literature is hardly a skill of mine. Are there any pro-seed oil persons claiming that linoleic acid doesn't make mice fat and ill? Or is that the sum total of "PUFAs vs mice" research?
I'm never going to believe that the whole field is corrupt or dishonest or lying. (Well, until I see it with my own eyes...). And I'm certainly not going to believe in conspiracies, because conspiracies require competence.
I am perfectly happy to believe that the whole field is a useless clusterfuck of subsidy-scrounging lamebrains who couldn't reason their way out of a paper bag, infested with politics and literally worse than useless, set up to be self defeating, producing random gibberish in vast quantities at state expense whilst not bothering to do basic sanity checks on their ideas. A fate worse than death that takes our brightest innumerate youngsters and trains them to slave away doing nothing worth doing, with their brains permanently disabled while they concentrate on looking and sounding wise, and explaining that everything is really complicated and difficult and that nothing can ever be truly known.
That's no problem for me, because we have the shining examples of the various humanities to prove that that is a thing that can happen. And before that, religion and theology and philosophy and alchemy.
And as far as I can see, none of the things that make the hard sciences truth-finding apply to medical science, so why would it be any different?
Indeed, I have suspected this to be the case for more than a decade now, since I first came into contact with the medical literature. Before that I was only a little sceptical.
But it does seem, in this particular case of "mice vs PUFAs", that I have somehow managed to hold these people in insufficient contempt. I am shame. Thank you for the correction.
I think you’re slightly misrepresenting that study tbh.
Yes, it appears that in an ad libitum (eating as desired) context, the low-fat group consumed fewer calories and consequently lost more weight and fat. However, if both groups were to consume the same number of calories, there might be a potential benefit to the low-carb group, especially in terms of fat loss.
The study results suggest that when individuals are allowed to eat without specific calorie restrictions, those following a low-fat diet tend to naturally consume fewer calories, which leads to greater weight and fat loss. On the other hand, the low-carb group, while not restricted in calorie intake, might have consumed more calories due to the nature of their diet, which includes foods that are often more calorie-dense.
If both groups were controlled to consume the same number of calories, it's possible that the low-carb group might have demonstrated a more significant advantage in fat loss, given the different metabolic responses associated with low-carb diets. However, this would be a hypothesis that would need to be tested through a controlled calorie-matched study.
Let's calculate the net benefit of fat loss in the low-fat group compared to the low-carb group, considering the difference in calorie intake.
Difference in Calorie Intake = Calorie Intake in the Low-Carb Group - Calorie Intake in the Low-Fat Group
= 2,584 calories/day - 2,243 calories/day
= 341 calories/day (not 700kcals, like you said).
Next:
Weight Loss Difference = Weight Loss in the Low-Carb Group - Weight Loss in the Low-Fat Group
= 9.6 kg - 11.6 kg
= -2 kg
Fat Loss Difference = Fat Loss in the Low-Carb Group - Fat Loss in the Low-Fat Group
= 3.4 kg - 4.8 kg
= -1.4 kg
Net Benefit of Fat Loss = Fat Loss Difference / Difference in Calorie Intake
= (-1.4 kg) / (341 calories/day)
≈ -0.0041 kg/calorie
The negative value indicates that for every additional calorie consumed by the low-carb group compared to the low-fat group, there was a net benefit of approximately 0.0041 kg of fat loss in favor of the low-carb group.
30-70% was from the titular graph of the study, which shows the low-carb line is significantly better than the low-fat group at any point in time, just to varying degrees.
> If both groups were controlled to consume the same number of calories
This would invalidate the study, because you're controlling for the result, aka fat loss. This malpractice is why Big Science only finds noise and no signal.
You can’t make such an inference from the graph, and you also can’t make the conclusion that if both groups were controlled to consume the same number of calories it would invalidate the results - or...you obviously can, but I fail to see the logic in such statements? Please elaborate.
I understand your concerns about data manipulation, and it's true that not all studies are perfect. However, the scientific method is designed to minimize biases and errors. It's important to consider that scientific consensus isn't based on a single study but on multiple studies that are peer-reviewed and replicated. Each study contributes to a bigger picture, helping us understand complex topics like nutrition better. While personal experiences are valuable, and I do read your posts and Brad Marshall’s content with interest even if a lot of it is contrary to accepted science - but be aware that they can't replace the comprehensive insights gained from large-scale research across the whole hierarchy of evidence. It's always good to approach studies critically, looking at their methodology and who's conducting them. I have no reason to believe that Kevin Hall is anything but honest and thorough in his data analysis and conclusions. I think we can make informed decisions based on a balance of personal experiences and scientific evidence, but absence of evidence doesn’t = evidence of absence when it comes to these quite creative mechanistic speculations that are often presented here and there, at least not when we have real-world outcome data across large populations and cohorts that are representative of the people we want to help.
How can I not? That's literally what the graph shows. At any given time, the LC diet group had lost 30-70% more weight than the LF group.
> you also can’t make the conclusion that if both groups were controlled to consume the same number of calories it would invalidate the results
This is literally controlling for (part of) the result. Weight loss == less calories retained. Of course it could also work by increasing the TDEE, which low-carb probably does, especially low-protein low-carb.
> It's important to consider that scientific consensus isn't based on a single study
The consensus seems to be based on dogma and group think from what I can see.
E.g. consider that the consensus is still that you should avoid saturated fat, for which there was never a shred of evidence, ever.
> I have no reason to believe that Kevin Hall is anything but honest and thorough in his data analysis and conclusions.
While there might have been periods where the low-carb group showed greater short-term effectiveness in weight loss, the final results indicated a different outcome favoring the low-fat group.
It's common to see variations in how different diets perform over time. Short-term results can sometimes be misleading, as they may reflect initial water loss, especially in low-carb diets, rather than true fat loss. The most relevant measure of a diet's effectiveness is its sustainable impact on weight loss over a longer period, which in this case, was demonstrated to be in favor of the low-fat group - and in an ad libitum context (which is, after all, how most people actually eat).
Nutritional science, inherently complex and dynamic, often appears dogmatic or contradictory. I’ve spent a career, and still a main part of my working day reviewing and evaluating research and I can assure you it’s both nuanced and evolving. This might not be immediately apparent, especially when confronted with selective, context-free studies (aka cherry-picking) often promoted by those with strong biases, like certain low-carb proponents in the online sphere. The scientific consensus is not static; it's based on repeated testing and critical reevaluation, continually updating when new research appears. The BCAA link may indeed prove to be relevant in certain contexts, but it’s still early to conclude just because there have been some promising anecdotes in a subreddit. Dismissing “Big Science” as confusing or misleading while resorting to personal anecdotes or online sources for conclusions is not just an intellectual oversight but could also lead to harmful health decisions. But sure, we can agree to disagree, I have no problems with that at all. I think you have some interesting perspectives and anything that makes me consider or reconsider evidence is welcomed 😊
> While there might have been periods where the low-carb group showed greater short-term effectiveness in weight loss, the final results indicated a different outcome favoring the low-fat group.
In the study I'm referring to, the low-carb group shows greater weight loss for every single time point measured, from 30-70% greater, hence my claim.
In a world with 43% obesity and 74% overweight rate, listening to Big Science is medically dangerous.
Kevin Hall's study, like many in nutritional science, aims to isolate specific factors to understand their impact on weight loss and metabolism. It’s a standard approach in scientific research to control variables to pinpoint the effects of interest. Regarding weight loss and Total Daily Energy Expenditure (TDEE), while low-carb diets might slightly increase TDEE, the effects can be modest and not have any real-world outcome effects. On a longer term basis, low-fat and low-carb diets are pretty much identical in free-living, ad libitum contexts - so individual preferences and adherence are the most likely factors to consider, not macro composition per se.
It's also important to note that the scientific consensus in nutrition evolves based on accumulating evidence, not through dogma or groupthink. Early studies on saturated fats did show a link to heart disease, which informed initial guidelines. However, recent research suggests the impact of saturated fats may depend on overall diet and individual factors, leading to a more nuanced understanding than we used to have. We can now isolate specific saturated fats, the food matrix plays a role (so butter is worse for a marker such as LDL whereas cheese is neutral), the overall fat intake etc. Saturated fat across studies of millions of people shows that below 10% of calories generally improves outcomes. The ones claiming that there’s no correlation move the breaking point and will compare e.g. 20%+ with below 20% - and when doing so it’s already including the range 10-20% where the health effects are exponentially modified, hence not stratifying properly. But I digress. I went deep into the whole LDL discussion with RationalDog on Reddit, which I’m sure you noticed, so I’m not going to repeat myself here.
> It's also important to note that the scientific consensus in nutrition evolves based on accumulating evidence, not through dogma or groupthink. Early studies on saturated fats did show a link to heart disease, which informed initial guidelines.
Sorry, but that's just false, both sentences.
It's well-documented that the initial saturated-fat studies were completely made up, and any critics were bullied and cancelled. This is well-documented in Big Fat Surprise.
And nutrition science is more dogmatic than the Catholic Church.
My money's on "it's the PUFAs only the Revolution will be impeded a ridiculous amount by Society." Idk man. It seemed implausible at first but I keep not seeing anything contradicting it. Fight the good fight!!!!
Well it's all trade-offs. Will the pre-cut chips keep longer? Will that save tiny amounts of money? Will anyone actually notice that it's killing people?
On the other hand, I do think that isoleucine itself is probably pretty harmless. It's the PUFAs that are the trouble.
The observation that mice fed diets high in polyunsaturated fatty acids (PUFA) versus saturated fats become obese, while humans are often advised to consume PUFA for health, certainly seems paradoxical at first glance. Let's break it down with a skeptical lens.
1. **Species Differences**: *Hmm, mice aren't tiny humans. Their physiology is different.* Mice and humans metabolize fats differently. What causes obesity in mice might not have the same effect in humans. For example, mice have a higher metabolic rate and different lipid processing mechanisms.
2. **Types of PUFA**: Not all PUFA are created equal. Omega-3 and omega-6 fatty acids are both PUFAs, but they have different effects on health. Omega-3s are generally considered anti-inflammatory and beneficial, while excessive omega-6s (which are more common in Western diets) can be pro-inflammatory.
3. **The Role of Diet Composition**: *It's not just about fat types, but the whole diet.* In studies, the context of the entire diet matters. Mice studies often use highly controlled diets, which don’t fully replicate the complexity of human diets. The impact of PUFA might differ in the context of a balanced diet containing fiber, protein, and other nutrients.
4. **Quantity and Quality**: The quantity of PUFA and the overall diet quality in humans matter. Excessive calorie intake, regardless of fat type, can lead to obesity. Also, processed foods high in PUFA may have other unhealthy components.
5. **Long-term vs Short-term Effects**: Mice studies often look at short-term effects, while human dietary guidelines consider long-term health outcomes. Short-term weight gain might not correlate with long-term disease risk.
6. **Generalizing from Animal Studies**: *Mice studies give clues, but they're not conclusive for humans.* Animal studies are a starting point for understanding biological mechanisms, but their findings need to be interpreted cautiously when applied to humans.
7. **Epidemiological Evidence in Humans**: Observational studies in humans have shown that diets high in certain types of PUFA (like omega-3s) are associated with reduced risk of heart diseases and other health benefits.
8. **Clinical Trials**: Controlled trials in humans provide more direct evidence. Many have shown benefits of PUFA-rich diets in reducing risk factors for heart disease.
In summary, while mouse studies are valuable, they can't be directly extrapolated to humans due to physiological differences. The current dietary recommendations for PUFA intake in humans are based on a large body of research, including epidemiological studies and clinical trials, which generally indicate health benefits, especially when replacing saturated fats with PUFA. *It’s a complex picture, and dietary advice must be nuanced.*
Hi Borge, please keep commenting here, you sound like a man who can show me when I'm missing important things, and that's my favourite kind of comment.
I very much agree that 'PUFAs are bad for mice' doesn't straightforwardly imply that 'PUFAs are bad for humans'.
In fact for nearly a year now, I *thought* that I was arguing that 'PUFAs are not bad for mice' didn't imply that 'PUFAs are not bad for humans'.
Which I was perfectly happy to do, mainly because I figured mice would be more tolerant of PUFAs than humans because of their natural diet.
But I *thought* that I had to explain why the PUFA thing might be true *in spite* of mouse studies.
Whereas in fact it seems that the mouse studies might be on my side.
And at that point, the whole 'seed oil hypothesis' just looks like a racing certainty to me. There's no mystery to explain, it can all just be straightforwardly true, and there's no real evidence against.
I'm shocked and surprised by this, and I'm still having trouble believing it. I keep thinking some pro-seed-oil person is going to leave me a comment with twenty studies proving that mice do fine on seed oils.
I would have thought that before people went around telling everyone about the benefits of polyunsaturated fats for seventy years or so, they would have at least checked to see whether they poisoned mice. That's the first thing we tend to do when testing to see whether things are safe to eat isn't it?
I wasn't expecting them to have done it well, but I was expecting them to have had a pop. It just never occurred to me that they hadn't bothered.
The progression from observing effects in animals to making dietary recommendations for humans isn't direct or simplistic. In the case of PUFAs, the shift towards incorporating them in human diets wasn't solely based on animal studies. It also involved epidemiological evidence in humans, observational studies, and controlled trials. Did researchers jump the gun? Perhaps not. These human studies generally showed that diets rich in certain types of PUFAs (especially omega-3 fatty acids) were associated with lower risks of heart disease and other health benefits.
Why Still Use Rodent Studies?
Rodent studies remain valuable for several reasons:
Basic Biological Insights: They offer insights into basic biological processes and mechanisms that might be relevant to humans.
Controlled Environment: Rodents can be studied in a controlled environment, allowing for the isolation of specific variables.
Ethical and Practical Reasons: It’s often not feasible or ethical to conduct certain types of experimental studies in humans.
Applicability of Rodent Studies to Humans....This is the tricky part. Not all mechanisms observed in rodents apply to humans.
The decision to consider a mechanism applicable depends on:
Biological Plausibility: Does the mechanism make sense in the context of human biology?
Consistency with Other Evidence: Does the observation align with epidemiological and clinical evidence in humans?
Further Research: Are there follow-up studies in humans that support or refute the findings from rodent studies?
The recommendation to include PUFAs in human diets wasn't primarily for obesity prevention but rather for overall cardiovascular health. While obesity is a risk factor for many diseases, the type of fat in the diet also plays a role in cardiovascular health, independent of obesity. It’s a balancing act. Recommendations for PUFA intake are part of a broader dietary context, emphasizing moderation, variety, and the replacement of less healthy fats.
Critique and Evolution of Dietary Guidelines: It's important to note that dietary guidelines evolve as new research emerges. Early recommendations may have been based on the best available evidence at the time, which has since been refined. Science is always a work in progress.
Indeed, not all PUFAs are the same. The two main types are omega-3 and omega-6 fatty acids. While both are essential fats that the body cannot produce on its own, they have different roles and effects:
Omega-3 Fatty Acids: Found in fatty fish, flaxseeds, chia seeds, and walnuts, these are generally considered anti-inflammatory and beneficial for heart health.
Omega-6 Fatty Acids: Common in many vegetable oils, nuts, and seeds, these are also essential but can be pro-inflammatory when consumed in excess, especially relative to omega-3s.
The context in which PUFAs are consumed is crucial. In ultra-processed foods, PUFAs are often present in refined oils and may be accompanied by high levels of sugar, salt, and unhealthy additives. The negative health impacts of ultra-processed foods could outweigh the benefits of PUFAs.
There's a growing understanding that the matrix in which nutrients are consumed matters. For instance, nuts and seeds contain fiber, vitamins, and other compounds that may synergistically contribute to their health benefits. In contrast, extracted oils are more concentrated in fat and lack these additional nutrients.
Comparing the fatty acids in cheese, milk, and butter is an important point – the overall dietary matrix. Recent research suggests that the health effects of dairy fats may depend on the other components of the dairy product and the overall dietary pattern where butter often elevates LDL and is considered risky when consumed in excess, whereas cheese, fermented dairy and milk is considered healthy/net positive.
While certain PUFAs are beneficial, they should be consumed as part of a varied diet rich in whole foods. Overconsumption of any type of fat, even healthy ones, can lead to negative health outcomes.
I think many are taking it too far when they obsess about low-PUFA pork or poultry if otherwise eating lean meats, and avoiding all nuts and seeds - when ALL relevant research considers these to be beneficial for a range of health effects. The average american consumed too many calories overall, where PUFAs are contained in the ultra processed junkfood they are eating. Not eating walnuts *because of PUFA* is missing the forest for the trees.
> Did researchers jump the gun?
Borge, thanks for everything you say here, it's all perfectly sound and true.
But do you seriously think it was wise to recommend putting PUFAs in food without checking to see whether they caused any problems in mice?
How would you react if some quantum holistic shyster recommended putting some aromatic compound derived from magic healing crystals in the food supply in very large quantities without even doing the most basic safety tests?
I honestly don't know the timeline of this, but again - mice aren’t small humans, many things that was bad for mice was neutral in humans, many things that were amazing in rodents were bad in humans etc etc.
> many things that was bad for mice was neutral in humans
Sure, ethanol springs to mind. But it still seems like the sort of thing that one might want to check?
But that’s what they did 🤷🏼 In many different trials.
The most extensive resource of human data I know of:
https://www.the-nutrivore.com/post/a-comprehensive-rebuttal-to-seed-oil-sophistry
Check out the Norwegian studies in the Tucker Goodrich link (https://tuckergoodrich.substack.com/p/does-linoleic-acid-induce-obesity). Maybe I'm wearing the omega-6 tinted glasses here, but it seems like pretty much exactly the study you're asking for - mice on 1% vs 8% LA, both on a high-fat diet (same total fat amount).
Maybe I'm being full of shit here, and reading something into it that isn't there. If so, please let me know.
I think the #1 reason that it's not an instant open/shut case is that it is indeed somewhat complicated:
1. There's a buffer/reservoir thing going on, so just stopping seed oils now doesn't guarantee you'll lose weight tomorrow
2. There's almost certainly a genetic or similar component, where it affects people to a different degree, though almost nobody seems to be entirely immune (<8% of U.S. population has 0 metabolic syndrome symptoms)
3. There is at least the interaction with BCAAs/isoleucine, maybe others. Again this doesn't seem to affect all people. It'd be interesting to see the individual data for the mice in those isoleucine studies - did they all respond the same, or did we have some lean mice and some super obese American mice? In fact, I think I'll ask Lamming if he has that data.
> Maybe I'm being full of shit here, and reading something into it that isn't there. If so, please let me know.
No, I'm beginning to think you're just straightforwardly right. But we need an intelligent enemy to tell us what we're missing. I wonder where we can find a clever pro-seed-oils man who'll give us the time of day?
Could try Stephan Guyenet? Maybe he'll answer your emails. He used to be anti seed oil (he turned Tucker Goodrich!) but then suddenly changed his mind for unknown reasons.
Yes, I've heard of him. If he used to think what I currently think and now doesn't I need to know what changed his mind. I wonder if he wrote it down anywhere?
I asked him once and I thought the answer was nonsense:
https://twitter.com/sguyenet/status/1679645193539309568
The rodent studies showing that 1% vs 8% LA content makes the difference with total fat content held same are referenced here:
https://tuckergoodrich.substack.com/p/does-linoleic-acid-induce-obesity
> It should be very obvious indeed, if it's true.
> A child should be able to do this experiment:
> Because if it's true, it should be known.
So adorable :)
Did you think this whole time that we were in a "race to find the truth" with Big Science? Of course not. Big Science has no interest in finding out what causes obesity; it'd ruin the whole racket.
It is trivial to test this. It has been tested. A child could do it.
And yet, it's been swept under the rug. When I talk to Actual Serious Scientists about it, they pretend they didn't understand the question or misheard me or radio silence.
I talked to Lamming, the guy who runs the lab with all those isoleucine studies, about this - hey, all your mice are on a high-PUFA diet, even the control mice have 3% LA and seem prediabetic from the glucose tolerance test!
He said he thinks his mice are only 1-2% LA. Turns out he didn't know how much LA is in corn oil, the main fat source in his control chow.
They wear CICO/fat tinted glasses. They cannot see the truth if you hit them over the head with it.
After over a decade on keto, following a bunch of interactions between ketoers and Big Science, I believe very strongly that a lot of mainstream scientists are straight up crooks. They're dishonest, they will lie, they will misrepresent their own findings. Maybe they believe their own lies, though it's hard to see how that could be possible.
Kevin Hall, THE big shot in U.S. nutrition science, likes to post a study of his titled something like "All diets work exactly the same," the title image (!) of which shows that low-carb diets work 30-70% better than low-fat diets.
He also published a study saying "high-fat ketogenic diet is obesogenic." The study: he compared a junk-food "high-fat" (=mixed) diet to a low-fat, practically vegan diet ("Whole Food Plant Based"). The high-fat people ate like 700kcal/day more (IIRC) yet lost the same amount of weight!
Physics defied! First law of Thermodynamics broken! CICO disproven! Right?!
No. Obviously, his finding was "Ketogenic diet is obesogenic by causing increased caloric intake."
He is simply a liar. Maybe he believes his own lies, which would make him a zealot, but there's no functional difference.
You seem to believe in the Efficient Market Hypothesis of Science - if it was easy to demonstrate, it would be common knowledge. I believe the Market for Science is completely rigged, and it is not at all efficient.
There are tons of these experiments. Grant from https://ggenereux.blog/ literally did the experiment you mentioned, just with vitamin A - he bought rodents, put them in cages, and didn't feed them any vitamin A. They never, ever got vitamin A deficiency. It's so trivial a child could do it, yet science "knows" you couldn't possibly survive for more than a few weeks without vitamin A.
I do actually think that if we find something that is reliable and easy to demonstrate, it won't just get ignored!
We have to actually do that though. At the moment, we've got you and me claiming to have lost considerable weight, and me claiming that my whack-job thyroid stuff is getting better, but neither of us are claiming to have got back to normal health and weight. In fact both of us are having real trouble. And that's about it.
I think that it all makes a certain theoretical sense, but even you don't take me seriously about all that. And you're right not to. Unless my theories end up showing results and predicting things, which at the moment they're not doing.
That's enough to get me interested, but it's not nearly enough that I could convince anyone sceptical that we'd found something important, or even to try it. Any fad diet has that kind of evidence, and they're mostly bollocks.
But if it becomes viral and popular, as it inevitably will if it actually works and we can convince people who are already open to trying things, then eventually a mountain of anecdotes will become real evidence, and then people will start to try to test it formally, and eventually it will win.
And one day, what is actually going on, right down to the molecular biology of it will be settled science.
Our job is to get it to work on us, first. And then if we can get it to work on us, to get it to work on other people.
Sure. It'll go like anything in this field (or exercise).
1. Some crazy practitioners try it
2. It's "crazy" and "dangerous" and "extreme"
3. If it actually works, plenty of people end up doing it anyway
4. The science swoops in to prove why they were always right about how it works
We're still at 1. Keto and carnivore are at 3.
I almost totally agree for once, except that it doesn't actually need to work to get to stage 3.
If we're right, then (no-PUFA) keto and carnivore will actually work, maybe real slowly, but the fact that they've got to stage 3 doesn't convince me that they work. It just sort of hints at it. And science is right to be very sceptical of them. You yourself got very fat on a keto diet!
I think it kinda, sorta needs to work to a degree.
Low-carb took off and it works in maybe 20-30% of people (worked better before the PUFA content increased so drastically, I imagine).
So did keto, which works in 30-50% of people, carnivore maybe 50-70%.
So we don't need to get to 100%, but it needs to work at least a little bit.
I did gain 100lbs on keto, but I also lost 100lbs on keto. It DOES work, just kind of orthogonally if you accidentally do it right ;) Carnivore has a higher hit ratio, because the only PUFA you can eat on it are chicken and pork, and it won't work if you're one of those low-protein people like me, but that seems to be maybe 30-50% of those who try it only.
Kate was telling me about her teenage experiences with the F-plan diet. Very popular at one point. Jesus Christ. Young women really will do just about anything to themselves to lose weight.....
What is the F-plan diet?
> You seem to believe in the Efficient Market Hypothesis of Science - if it was easy to demonstrate, it would be common knowledge. I believe the Market for Science is completely rigged, and it is not at all efficient.
Depends on the science.
In fundamental physics, I do believe EMHS pretty much. If you can find something that routinely seems to violate the currently understood laws of physics (which we know are wrong!), then physicists will be very very interested, and they will try to pin down the precise mechanisms, and if necessary, they will upend their whole world view to account for the tiny detail.
In English Literature, say, there are no laws, anything goes, nobody cares. They don't make any claims that could be tested, they don't even believe in truth. I see no reason at all to trust them about anything, although I can sometimes see a certain value to some of their work.
The question is:
Where on the spectrum of all the things that get done in universities does medical stuff lie?
And I think that things like molecular biology are trying very hard to be the first type of thing, and having great success.
And I think that things like nutrition advice and psychology seem more like the second type of thing.
They're getting better though. People take the 'replication crisis' very seriously, and are trying to reform the system so it more reliably leads to finding out true things. But it's really hard. Managing large scale collections of people is always really hard. Changing systems is really hard.
Yea it's definitely closer to the English Lit Major end of the scale.
I actually think it's gotten much, much worse. Our state of knowledge in nutrition is significantly worse than it was 100 years ago - we now un-know several true things, and we know a lot of falsehoods.
> I actually think it's gotten much, much worse.
Yes, there's quite a lot of that sort of thing around in lots of fields. That's what happens if your methods aren't self-correcting.
By the way, last night I was lying awake thinking about mice (as one does) and it dawned on me: of course there's a somewhat mainstream (in carnivore/keto circles) explanation for why even the low-PUFA, "high-fat" mixed diet causes obesity: The Randle Cycle. The Swamp.
These diets are typically 45% fat, and therefore still high in carbs. The "chow" diets are often 10% fat, rest mostly carbs. So a fair comparison would be a ketogenic 10% carb diet.
This still doesn't explain why the Norwegians managed to keep their mice lean on a 1% LA, 60% fat diet.. is 60% fat already enough in absence of LA? Or is there something else we're missing?
> Did you think this whole time that we were in a "race to find the truth" with Big Science?
Yes, that is what I thought. The only reason I thought we had a chance is that they are so useless.
It is what I still think. If we've come up with a simple cheap technique that works to fix obesity and all the other diseases of modernity (which I *can* believe, but do not yet believe), then it will spread like wildfire, and there is nothing anyone is going to be able to do to stop it.
If it all turns out to work then you, I think, were the one that made the crucial observation. No way on earth will you ever get the credit for it, but we will know. And I can feel good about having been able to spot the one man talking sense in all the gibberish.
Do you think it's a coincidence that tens of thousands of professional scientists funded by taxpayer money apparently haven't figured out what a pig farmer from upstate New York and some asshole like me and a boat-dwelling, choir-singing, chips-frying bon vivant from Cambridge managed to figure out in a handful of years?
They aren't even looking for it. Of course there are a tiny % that are doing helpful stuff and go unnoticed, like Lamming. But it's rotten at the top, much of the way down, I think.
edit: And yes, if it works good enough (70%?), it might spread and take over and solve the issue (maybe to 70% or something). The reason it hasn't is because this is only true as of social media.
When I first heard of keto, which was already 20 years or so into the internet, it was that dangerous evil diet nobody could possibly survive, and which would almost certainly kill you. Carnivore wasn't even a thing people could fathom.
> Do you think it's a coincidence that tens of thousands of professional scientists funded by taxpayer money apparently haven't figured out what a pig farmer from upstate New York and some asshole like me and a boat-dwelling, choir-singing, chips-frying bon vivant from Cambridge managed to figure out in a handful of years?
Well how many of them were actually thinking about the problem?
The existence of standard breeds of lab mice and standard mouse diets shows that they're trying to be careful.
I'm interested in "PUFAs bad" on the basis of evolutionary thinking that makes sense to me, but that no one else takes seriously at all, and in ex150 originally because your original arguments made intuitive sense to me, and now also because it does seem to work well for me.
Brad is interested in "PUFAs bad" because of molecular biology that I don't understand nearly well enough to evaluate, and on the basis of torpor/hibernation theories that I think are risible. And why he's suddenly got interested in BCAAs is beyond me, but it rhymes quite nicely with the protein restriction bit of ex150.
If we can raise "PUFAs bad, protein involved" to the level of a hypothesis that people take seriously at all, then suddenly all those diets will get their various lipid levels measured and controlled very accurately, and then they'll have the tools to actually discover stuff.
> Well how many of them were actually thinking about the problem?
It is kind of their job, isn't it.
Sorry, I mean, "Thinking about the problem through the lens of PUFAs bad". There hasn't been anything to raise it to their attention out of the utter swamp of plausible competing ideas.
I've suspected all my life that saturated fats and salt are fine, on exactly the evolutionary grounds that you think are dangerously misleading.
I've wondered whether PUFAs are actually actively harmful for about the last year. There was nothing to raise it to my attention either, and I'm unusually interested in crackpottery in all its forms.
They've resisted the idea very vigorously, and they lied & cheated & bullied to push it into the existing paradigm. It's not a coincidence.
Read Big Fat Surprise, this is all well-documented. Mere incompetence is not sufficient to explain it.
> The rodent studies showing that 1% vs 8% LA content makes the difference with total fat content held same are referenced here:
> https://tuckergoodrich.substack.com/p/does-linoleic-acid-induce-obesity
Nice, thanks for that. I found:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5235953/
It seems to me that that's it? But trawling through the medical literature is hardly a skill of mine. Are there any pro-seed oil persons claiming that linoleic acid doesn't make mice fat and ill? Or is that the sum total of "PUFAs vs mice" research?
I'm not super aware of the entire literature either, so you'd have to ask a pro-seed oil person.
This "Kevin Hall" study sounds hilarious. Can you bung me a link? I need a bit of light reading.
I think this is one of the 2: https://www.nature.com/articles/s41591-020-01209-1
Can't find the other one right now
I'm never going to believe that the whole field is corrupt or dishonest or lying. (Well, until I see it with my own eyes...). And I'm certainly not going to believe in conspiracies, because conspiracies require competence.
I am perfectly happy to believe that the whole field is a useless clusterfuck of subsidy-scrounging lamebrains who couldn't reason their way out of a paper bag, infested with politics and literally worse than useless, set up to be self defeating, producing random gibberish in vast quantities at state expense whilst not bothering to do basic sanity checks on their ideas. A fate worse than death that takes our brightest innumerate youngsters and trains them to slave away doing nothing worth doing, with their brains permanently disabled while they concentrate on looking and sounding wise, and explaining that everything is really complicated and difficult and that nothing can ever be truly known.
That's no problem for me, because we have the shining examples of the various humanities to prove that that is a thing that can happen. And before that, religion and theology and philosophy and alchemy.
And as far as I can see, none of the things that make the hard sciences truth-finding apply to medical science, so why would it be any different?
Indeed, I have suspected this to be the case for more than a decade now, since I first came into contact with the medical literature. Before that I was only a little sceptical.
But it does seem, in this particular case of "mice vs PUFAs", that I have somehow managed to hold these people in insufficient contempt. I am shame. Thank you for the correction.
> I'm never going to believe that the whole field is corrupt or dishonest or lying. (Well, until I see it with my own eyes...)
We'll, I'm way ahead of you there - I've seen it with my own eyes.
I think you’re slightly misrepresenting that study tbh.
Yes, it appears that in an ad libitum (eating as desired) context, the low-fat group consumed fewer calories and consequently lost more weight and fat. However, if both groups were to consume the same number of calories, there might be a potential benefit to the low-carb group, especially in terms of fat loss.
The study results suggest that when individuals are allowed to eat without specific calorie restrictions, those following a low-fat diet tend to naturally consume fewer calories, which leads to greater weight and fat loss. On the other hand, the low-carb group, while not restricted in calorie intake, might have consumed more calories due to the nature of their diet, which includes foods that are often more calorie-dense.
If both groups were controlled to consume the same number of calories, it's possible that the low-carb group might have demonstrated a more significant advantage in fat loss, given the different metabolic responses associated with low-carb diets. However, this would be a hypothesis that would need to be tested through a controlled calorie-matched study.
Let's calculate the net benefit of fat loss in the low-fat group compared to the low-carb group, considering the difference in calorie intake.
Difference in Calorie Intake = Calorie Intake in the Low-Carb Group - Calorie Intake in the Low-Fat Group
= 2,584 calories/day - 2,243 calories/day
= 341 calories/day (not 700kcals, like you said).
Next:
Weight Loss Difference = Weight Loss in the Low-Carb Group - Weight Loss in the Low-Fat Group
= 9.6 kg - 11.6 kg
= -2 kg
Fat Loss Difference = Fat Loss in the Low-Carb Group - Fat Loss in the Low-Fat Group
= 3.4 kg - 4.8 kg
= -1.4 kg
Net Benefit of Fat Loss = Fat Loss Difference / Difference in Calorie Intake
= (-1.4 kg) / (341 calories/day)
≈ -0.0041 kg/calorie
The negative value indicates that for every additional calorie consumed by the low-carb group compared to the low-fat group, there was a net benefit of approximately 0.0041 kg of fat loss in favor of the low-carb group.
Where did you come up with the 30-70% figure?
30-70% was from the titular graph of the study, which shows the low-carb line is significantly better than the low-fat group at any point in time, just to varying degrees.
> If both groups were controlled to consume the same number of calories
This would invalidate the study, because you're controlling for the result, aka fat loss. This malpractice is why Big Science only finds noise and no signal.
You can’t make such an inference from the graph, and you also can’t make the conclusion that if both groups were controlled to consume the same number of calories it would invalidate the results - or...you obviously can, but I fail to see the logic in such statements? Please elaborate.
I understand your concerns about data manipulation, and it's true that not all studies are perfect. However, the scientific method is designed to minimize biases and errors. It's important to consider that scientific consensus isn't based on a single study but on multiple studies that are peer-reviewed and replicated. Each study contributes to a bigger picture, helping us understand complex topics like nutrition better. While personal experiences are valuable, and I do read your posts and Brad Marshall’s content with interest even if a lot of it is contrary to accepted science - but be aware that they can't replace the comprehensive insights gained from large-scale research across the whole hierarchy of evidence. It's always good to approach studies critically, looking at their methodology and who's conducting them. I have no reason to believe that Kevin Hall is anything but honest and thorough in his data analysis and conclusions. I think we can make informed decisions based on a balance of personal experiences and scientific evidence, but absence of evidence doesn’t = evidence of absence when it comes to these quite creative mechanistic speculations that are often presented here and there, at least not when we have real-world outcome data across large populations and cohorts that are representative of the people we want to help.
Can I just say how happy I am to see an argument breaking out in the comments section of my blog that doesn't involve me! It feels like a milestone.
Have at it, you two, I am interested to find out what you eventually conclude.
How can I not? That's literally what the graph shows. At any given time, the LC diet group had lost 30-70% more weight than the LF group.
> you also can’t make the conclusion that if both groups were controlled to consume the same number of calories it would invalidate the results
This is literally controlling for (part of) the result. Weight loss == less calories retained. Of course it could also work by increasing the TDEE, which low-carb probably does, especially low-protein low-carb.
> It's important to consider that scientific consensus isn't based on a single study
The consensus seems to be based on dogma and group think from what I can see.
E.g. consider that the consensus is still that you should avoid saturated fat, for which there was never a shred of evidence, ever.
> I have no reason to believe that Kevin Hall is anything but honest and thorough in his data analysis and conclusions.
Agree to disagree then.
While there might have been periods where the low-carb group showed greater short-term effectiveness in weight loss, the final results indicated a different outcome favoring the low-fat group.
It's common to see variations in how different diets perform over time. Short-term results can sometimes be misleading, as they may reflect initial water loss, especially in low-carb diets, rather than true fat loss. The most relevant measure of a diet's effectiveness is its sustainable impact on weight loss over a longer period, which in this case, was demonstrated to be in favor of the low-fat group - and in an ad libitum context (which is, after all, how most people actually eat).
Nutritional science, inherently complex and dynamic, often appears dogmatic or contradictory. I’ve spent a career, and still a main part of my working day reviewing and evaluating research and I can assure you it’s both nuanced and evolving. This might not be immediately apparent, especially when confronted with selective, context-free studies (aka cherry-picking) often promoted by those with strong biases, like certain low-carb proponents in the online sphere. The scientific consensus is not static; it's based on repeated testing and critical reevaluation, continually updating when new research appears. The BCAA link may indeed prove to be relevant in certain contexts, but it’s still early to conclude just because there have been some promising anecdotes in a subreddit. Dismissing “Big Science” as confusing or misleading while resorting to personal anecdotes or online sources for conclusions is not just an intellectual oversight but could also lead to harmful health decisions. But sure, we can agree to disagree, I have no problems with that at all. I think you have some interesting perspectives and anything that makes me consider or reconsider evidence is welcomed 😊
> While there might have been periods where the low-carb group showed greater short-term effectiveness in weight loss, the final results indicated a different outcome favoring the low-fat group.
In the study I'm referring to, the low-carb group shows greater weight loss for every single time point measured, from 30-70% greater, hence my claim.
In a world with 43% obesity and 74% overweight rate, listening to Big Science is medically dangerous.
Kevin Hall's study, like many in nutritional science, aims to isolate specific factors to understand their impact on weight loss and metabolism. It’s a standard approach in scientific research to control variables to pinpoint the effects of interest. Regarding weight loss and Total Daily Energy Expenditure (TDEE), while low-carb diets might slightly increase TDEE, the effects can be modest and not have any real-world outcome effects. On a longer term basis, low-fat and low-carb diets are pretty much identical in free-living, ad libitum contexts - so individual preferences and adherence are the most likely factors to consider, not macro composition per se.
It's also important to note that the scientific consensus in nutrition evolves based on accumulating evidence, not through dogma or groupthink. Early studies on saturated fats did show a link to heart disease, which informed initial guidelines. However, recent research suggests the impact of saturated fats may depend on overall diet and individual factors, leading to a more nuanced understanding than we used to have. We can now isolate specific saturated fats, the food matrix plays a role (so butter is worse for a marker such as LDL whereas cheese is neutral), the overall fat intake etc. Saturated fat across studies of millions of people shows that below 10% of calories generally improves outcomes. The ones claiming that there’s no correlation move the breaking point and will compare e.g. 20%+ with below 20% - and when doing so it’s already including the range 10-20% where the health effects are exponentially modified, hence not stratifying properly. But I digress. I went deep into the whole LDL discussion with RationalDog on Reddit, which I’m sure you noticed, so I’m not going to repeat myself here.
> It's also important to note that the scientific consensus in nutrition evolves based on accumulating evidence, not through dogma or groupthink. Early studies on saturated fats did show a link to heart disease, which informed initial guidelines.
Sorry, but that's just false, both sentences.
It's well-documented that the initial saturated-fat studies were completely made up, and any critics were bullied and cancelled. This is well-documented in Big Fat Surprise.
And nutrition science is more dogmatic than the Catholic Church.
My money's on "it's the PUFAs only the Revolution will be impeded a ridiculous amount by Society." Idk man. It seemed implausible at first but I keep not seeing anything contradicting it. Fight the good fight!!!!
Doing a quick google of the content of isoleucine in potatoes revealed this delightful study: https://www.sciencedirect.com/science/article/abs/pii/S0925521421003112
Will humanity insist on putting every foot wrong until it’s too late?
Well it's all trade-offs. Will the pre-cut chips keep longer? Will that save tiny amounts of money? Will anyone actually notice that it's killing people?
On the other hand, I do think that isoleucine itself is probably pretty harmless. It's the PUFAs that are the trouble.
Yes you’re right. Once they chuck out the pufa-laden chips it won’t matter if they use isoleucine in small quantities (maybe) on them.
> Once they chuck out the pufa-laden chips it won’t matter if they use isoleucine in small quantities (maybe) on them.
With any luck! I do hope someone remembers to check that in mice.