In Short
Excess linoleic acid from seed oils is the root cause. This makes its way up the food chain, including many animal products and nearly all packaged foods.
Given this high-linoleic environment, certain other dietary strategies can become conditionally problematic:
Mixing carbs and fats in your diet (“swamping”)
Eating moderate/high protein
The solution therefore looks like this:
Cut down linoleic acid aggressively, as close to ancestral levels of ~2% kcals of total TEE as possible, forever
In the likely case this doesn’t induce rapid fat loss, temporarily (=for a few years) do one or both of the following interventions:
Pick either a keto or “carbo” (high-carb, low-fat) diet
Additionally, restrict protein to <10% or even 6% of TEE
Introduction
If you’ve been reading this blog for a while, you’ve probably got a pretty decent understanding of my (current) mental model of obesity, and how to reverse it. You might even remember how I learned about Modern PUFA Theory and started incorporating it into my ideas, picking up ideas from Brad Marshall, Tucker Goodrich, and the fine folks at r/SaturatedFat.
I thought I’d write my current mental model down more explicitly. Not only as an easy overview for people who haven’t been following along, but also because writing my thoughts up helps me solidify them and find errors, inconsistencies, and ambiguities.
This model hasn’t changed dramatically in a while, but it is open to change. I don’t think this covers 100% of the modern obesity epidemic, but probably about 85% or so.
Meaning, I think these causes are behind 85% of modern obesity, and 85% of modern obese people could reverse their obesity by following the recommendations.
This is not about Abs
I don’t care very much for abs. If you look at pictures of healthy ancestral populations, be it hunter-gatherers or farmers, they rarely look like modern, shredded bodybuilders. The men are maybe around 18-25% body fat, the woman about 10% higher.
My goal is to find a way for people to eat to appetite and be healthy.
If you want to be on the cover of Men’s Health, you might well have to do extra things, like starve yourself down. Bodybuilders infamously starve themselves down to single digit body fat percentages, and it takes a massive toll on their health.
Reversing obesity and reaching a healthy weight is very different from getting shredded.
When you talk to the influencers on the internet trying to sell you coaching or other methods or “diet secrets,” they typically eventually admit that they can only maintain their abs by severely restricting their total caloric intake in perpetuity. Often times, these very active, adult men who allegedly train in the gym 5x/wk will eat only 2,000kcal, objectively the TEE of a young teenager or a very small woman. “You’re always hungry,” some will admit.
They are more dishonest than the bodybuilders. At least, bodybuilders will openly admit that they are only starving themselves down to peak shape for a single day photo shoot or bodybuilding show.
The lifestyle gurus are trying to sell you a dream: that everyone can run around with Men’s Health cover page abs all year long, and all you have to do is give them your money.
This is all well documented and I’m not interested.
Reversing Obesity
45% of Americans are currently obese, while 75% are overweight, which includes the obese.
My goal is to find a relatively straightforward way for these 45% of obese people to get down to non-obesity or close without going to bed hungry, using willpower to fight their own biochemistry, and all that nonsense that the standard dieting approach will tell you to do.
I’ve done it myself:
I went from nearly morbidly obese (which is just over 300lbs for me) to below the obesity line (“merely overweight”) in about 16 months, and I’ve stayed down there for most of a year since.
The only time I spent significant time above that line was a very recent dietary experiment eating only rice for a month.
A few weeks after ending the rice experiment, I’m now generally just below or at that line again.
In short, it was pretty easy, using my ex150 diet, to reverse my obesity and keep it off.
This involved no restricting caloric intake, no willpower. I ate to satiety every day. As far as I can tell (via the DEXAs I have) I lost no lean mass. I have no loose skin. My metabolic rate, as measured by RMR & DLW, is exactly where it should be, so I didn’t “tank my metabolism” - and why would I, I didn’t starve myself ever.
All the fat came off because, through dietary changes, my body suddenly got access to all the excess adipose tissue, and simply used it for fuel.
Non-Solutions I’m not Interested In
Starving yourself down temporarily
Yea, you can starve yourself in the short-term. And then you’ll gain all the fat back when you eat ad-libitum again. I’ve done it many times, and I’ve done it once during the last 2 year period for fasting experiments.
I dropped to 210lbs, by far the lowest I’d weighed in years.
And, of course, it all came back within 3 weeks when I went back to ad-libitum eating.
Not interested.
Wrecking your metabolism in the process
If you reverse your obesity but tank your metabolism in the process, lowering your metabolic rate significantly below where it should be, that probably means your progress is temporary - and that’s not a trade I would make.
If you’re obese, there’s no reason your metabolism should decrease markedly when losing the excess weight. After all, your body is carrying around immense reserves of fat that you should be burning off. No need to ever catabolize lean mass for energy.
A marked reduction in your metabolic rate is a sign that something’s not working as it should.
Losing a lot of lean mass
Similar to the above, you shouldn’t lose significant amounts of lean mass, which would, incidentally, lower your metabolic rate.
You’ll likely lose a little bit of excess skin, and possibly a small amount of muscle or other supporting lean mass no longer needed to maintain all that excess fat or weight.
But some of the GLP-1 drugs seem to cause 30-40% lean mass loss, which is crazy to me. Even the 25% lean mass loss that are considered “normal” when losing weight via caloric restriction seem extremely high to me.
I’d like to lose just about 0% lean mass when reversing obesity.
True, this might not be possible when starving yourself down to single-digit body fat levels. But then again, that’s not my goal.
Having to Restrict Calories Forever
If your “solution” relies on going hungry and eating the TEE of a little girl forever, it’s not a solution, it’s a “hack.” It can temporarily work, but it sure isn’t fixing any of the root causes of the obesity epidemic.
Almost nobody restricted their food intake 100 years ago, and yet, somehow, almost nobody was obese.
Infinite Complexity
Some people like to throw out “THAT’S TOO SIMPLE! IT’S MORE COMPLICATED!” and yes, sometimes that’s true.
But if we look anecdotally, it doesn’t actually seem all that complicated.
I believe that obesity is only slightly complicated.
There are about 8 billion people out there. But you can apparently cover about 85% of their obesity problems with a small handful of diets. We don’t need 8 billion individual solutions, only maybe 2-3 with a few modifications. Heck, make it 5.
Slightly complicated means there isn’t 1 exact diet that works for everyone, but we also definitely don’t need a new, specific diet for every single individual out there.
Slightly complicated means 3-5 solutions for 85% of the people.
With all that said, let’s now get into my current understanding of what caused the obesity epidemic.
Root cause: seed oils caused the Obesity Epidemic
It’s just the best fit in terms of timeline and epidemiology and explains the Mysteries of Obesity better than any other explanation I’m aware of.
Modern, industrial seed oils were invented around 1860 and have since made their way into every aspect of the food supply. This lines up better than anything else with the timeline of the obesity epidemic.
We also have several mechanisms for how linoleic acid, the “bad fat” in seed oils, causes you to overeat, shuts down your satiety mechanism, and causes several other bad things.
The linoleic acid from seed oils filters up the food chain as we feed it to chicken & pork, and PUFAs now make up about 20-25% of the Standard American Diet by carolies.
Many “healthy” options like the dressings on salads are made with seed oils, and while most “ultra-processed junk foods” are full of them, you can’t simply avoid them by “eating whole foods” and “cooking your own food” - which explains why these strategies often fail to work for so many people.
The solution is “simple” but very difficult: cut out as much linoleic acid from your diet as possible to get close to the pre-modern levels our bodies evolved to deal with.
This is difficult because linoleic acid is in everything; it accumulates up the food chain and pretty much all modern animals will be fed a high-LA diet and their fats will be too high in it.
This means either eating only extremely lean meat (pork/chicken) or sticking to ruminants, who can convert the dietary linoleic acid into the healthier fatty acids via the bacteria in their stomachs.
In short, when it comes to animal fats: stick to beef, dairy & lamb.
It also means cutting out almost all modern convenience foods, including “health” foods such as most nuts & salad dressing (almost always made with seed oils), sauces (ditto), and of course nearly all fried foods (almost always fried in seed oils). Most breads contain seed oils, as do pastries and nearly all prepared dishes.
Almost all restaurant foods are cooked in seed oils by default, including “butter blends” and “olive oil blends” (usually 1% butter/olive oil and 99% seed oil). I personally don’t bother asking the waitstaff what they cook with and don’t ask to have my food cooked in butter/olive oil, as most of them don’t know. Even if they pass on your directions, it’s unlikely the busy line cooks will have time or the real fats on hand.
It’s safer to just not eat out at restaurants.
Maybe consumer preferences will change the market over time. There seem to be plenty of companies switching away from seed-oils for health reasons. But until that becomes more widespread, it’s just safer not to take your chances.
There are some notable exceptions that I make use of when traveling:
McDonald’s burger patties are pure beef, cooked in their own juices. They are widely available and safe.
Dairy is safe, and Starbucks can make you “breve” (=half & half) or heavy whipping cream lattes at pretty much any location country wide, including airports.
Dark chocolate is very low in linoleic acid and widely available, including at travel locations like airports.
(Note that these are more suited for high-fat diets. I’ve been doing a high-fat/low-carb/low-protein diet, but there are also high-carb/low-fat alternatives out there. You just have to find some safe staples for when you can’t cook for yourself, e.g. while traveling.)
Downside: depleting PUFAs takes years
This is a huge downside both in terms of testing/explaining the veracity of the hypothesis, and also for individuals looking to go on a low-PUFA diet.
We know that the fats you eat get stored in your body fat and used in building your tissues, and that the process of switching out all fatty acids in your adipose completely can take 4-8 years.
That’s why “just cutting out PUFAs” doesn’t typically result in dramatic fat loss except for those who never had very much to lose in the first place.
People do typically report lots of small benefits when cutting out PUFAs, and even a return to “normal appetite” is common. But since a lot of PUFAs are still stuck in your body and get released every day, it’s sort of like you’ve put yourself on a seed oil drip for years to come.
That’s why most of the people who are seriously overweight need to also implement one of several medium-term interventions. Anecdotally, medium-term here means at least a year or so.
It’s not entirely clear yet that everyone who’s been severely obese will be able to stop doing these interventions eventually. We have some promising anecdotes of people who were able to.
Also, anecdotally, there seem to be historic populations that did not implement either of these interventions and were in much better metabolic shape than us. For example, my grandparents never did low-carb or low-fat, and while they didn’t exactly eat carnivore, they also didn’t consciously restrict protein intake.
That said, while it would be nice, it’s not clear that cutting out PUFAs for long enough will make these interventions unnecessary for everyone.
Time will tell.
Conditional intervention 1: Pick keto or carbo
One observation that many have made, but that was maybe formulated best by Denise Minger, is the parable of the Metabolic Swamp and that you want to avoid it.
There are plenty of people on “stupid internet diets” that seem to effortlessly lose lots of fat. These people tend to do very low carb diets or very low fat diets.
They rarely mix the two.
This simple observation about “avoiding the metabolic swamp” of mixing carbs & fats is further substantiated by looking at an overview of some ancestral cultures and their diets that we know of.
I’ve previously shown this in my Metabolic Swamp visualizer:
The white dots represents known dietary macros from various peoples, ancestral & modern.
You can see that only a single entry is directly in the “Swamp” area, and that’s the Standard American Diet.
The vast majority of recorded ancestral diets are in the “low-fat” or “carbo” zones. These are typically cultures eating lots of starches like rice or sweet potatoes, sometimes honey like the Hadza.
There are a few entries in the “keto” or “low-carb” area, but the only “ancestral” ones over there are certain Eskimo populations and the arctic explorer Steffanson, who replicated one such Eskimo diet living in New York City in the 1920s.
This is of course confounded; we don’t have very many direct records of dietary macros for all peoples who ever lived. Most didn’t leave detailed records, and since obesity was usually not a big issue, they certainly didn’t leave detailed records of their macros or food intake.
More recent cultures have left more records, and we know that the agricultural revolution happened around 10,000 years ago. It’s therefore hard to say what exact macros people were eating in the millions of years before that. There are very few if any records from before the agricultural revolution. Isotope studies can tell us if people ate “animal products” or “plant products” but are mostly focused on the protein content, and wouldn’t give us an exact macro breakdown.
I’ll therefore say it is up in the air if “avoiding the swamp” is a temporary intervention or if you have to do it forever.
Like I mentioned, there are some anecdotes of people who, after years of avoiding PUFAs enough, were able to swamp without gaining weight again.
But it’s far from an open & shut case at this point.
How to modify keto/carbo
Luckily there are lots of people out there doing either of these diets. You’ll have to take their recipes & thoughts with a grain of salt, because linoleic acid restriction is a key part that neither of these diet tribes will typically prioritize. Still, you’ll be able to find lots of recipes & ideas.
Keto
You can either do a ketogenic diet or even carnivore. Note that carnivore doesn’t work as well with protein restriction, the other conditional intervention. It can be done, but is quite tricky.
Your main point of contention with most keto dieters will be the very high content of linoleic acid in what I call “Standard American Keto,” which is quite high in bacon, roasted chicken, salad dressings, and nuts.
You should stick to beef only, or potentially super lean pork/chicken. Personally, I’d rather eat fatty beef than lean pork/chicken anyway.
You’ll have a much harder time eating out at restaurants & eating convenience foods: there is now a huge market for “keto” products & meals out there, but many of them contain terrible high-PUFA franken-ingredients or are entirely made of soybean oil (salad dressings).
Carnivore
Carnivore is actually remarkably close to a low-linoleic acid diet if you just stay away from bacon/pork/chicken. Typical carnivores prioritize beef, which is a safe animal product even when very fatty.
They also tend to avoid the soybean oil-laden sauces & dressings that can sneak in on keto, as well as seeds & nuts, another big source of linoleic acid.
Just be aware that many carnivores promote bacon & roast chicken, which I’d stay away from. You can get beef bacon, which tastes great, if you must.
Carbo
“Carbo” isn’t a common term, it’s just more convenient than all the other names. On this side of the spectrum, you can follow people practicing veganism, vegetarianism, “high-carb/low-fat” diets, Whole Food Plant Based diets, and more.
The main point of contention is, again, linoleic acid. Many of these diets tend to be low-fat overall, which makes it relatively easy to make them low-linoleic acid.
Just be aware that most liquid plant oils aren’t your friend, and check that the staples aren’t high in PUFA: some vegan staples like quinoa or nuts are super high in linoleic acid. I wouldn’t use olive oil, personally, as it’s still 8-20% linoleic acid even when not adulterated.
You also don’t have to be as dogmatic about animal products, which makes these diets much easier to comply with and, in my humble opinion, much more healthy & sustainable long term.
There’s clearly something unhealthy about most long-term vegans. But you don’t have to be fully vegan, you only need to be “high-carb/low-fat!”
By eating a small portion of beef a day, you’ll avoid pretty much all long-term health issues that veganism presents: you’ll get your B vitamins in, you’ll have a high-quality protein source, you’ll get enough creatine, you’ll get some healthy, high-quality fats in to prevent skin & hormonal issues.
The diet will also be much less restrictive in that sense; while you have to pay extra attention to the linoleic acid content, you’re not limited to purely plant based foods, which opens up a lot of options.
Conditional intervention 2: restrict protein
Even doing a very strict ketogenic diet for years, I regained 100lbs. I eventually ended up brute-forcing my way to a highly protein-restricted diet and the fat just began to melt off.
Imagine my surprise when I found out that there’s a whole bunch of research demonstrating that protein restriction can inhibit & reverse the bad effects of the “obesogenic Western Diet” in lab animals.
In addition, protein restriction seems a common theme among the “stupid internet diets” that make people miraculously lose fat.
My own version, ex150, is only about 6% protein (and 4% carbs) and is therefore the extreme ketogenic version of a protein-restricted diet.
But several of the prototypical high-carb/low-fat diets are also quite low in protein. The infamous Kempner Rice Diet, which Denise Minger talks about in the aforementioned post on the Metabolic Swamp, is even lower in protein at 4%.
The Potato Diet, which also has achieved considerable internet fame, is also relatively low in protein depending on how you do it: a diet consisting of 100% potatoes would only be about 7% protein.
If we take another look at our handy swamp map, we can also see that almost no culture historically ate more than 20% protein. The Standard American Diet, at 16% protein, is actually on the higher-end of protein content historically.
Anecdotally, if we look at what our grandparents ate, it certainly wasn’t a high-protein diet in the modern broscience sense.
They were probably eating meat daily or almost daily, but in relatively small portions. It wasn’t a steak for every meal.
On the other hand, they also weren’t exactly restricting protein consciously: it was just relatively expensive compared to carbs and fat to buy high-quality animal protein.
Let’s call it a moderate protein diet.
The conditionality of protein consumption is as open as swamping: there are some anecdotes, but no clear trend yet.
Maybe, if we all avoid linoleic acid for long enough, we can all go on a high-protein diet and be totally fine. Maybe some of us need to stick to moderate-protein diets, and maybe some basket cases like myself need to stay at 6% forever.
The Gauntlet: Try out which interventions you need personally
I’d love to tell you there’s some test you can take, and maybe there will be one. So far, I’m not aware of any way better than simply trying it.
If you only ever were 20-30lbs overweight, maybe just cutting out the PUFAs is enough.
If you’re morbidly obese, or very obese like I was, it’s likely you’ll need to at least stick to one side of the metabolic swamp.
If you’re “lucky,” you might also have to restrict protein - I needed to. Some people don’t need to.
“Gauntlet” was an early video game and spawned a whole genre: heroes fighting their way through monster-filled dungeons, room by room.
I think that “Running The Gauntlet” (also often used in militaries or other functions) is an apt metaphor here. Most people who have reversed significant obesity and lost lots of weight have tried a lot of diets.
It’s rare for someone to just try a diet and “get lucky” on the first attempt. I view my own journey as a series of 30-day experiments & challenges, some of which help me lose fat, and many of which don’t.
It’s very similar to fighting your way through a dungeon, one room/experiment at a time, trying to avoid missteps and find your way out (of obesity).
I therefore suggest just doing one 30 day trial at a time:
Start by aggressively eliminating linoleic acid for 30 days. This should make you feel much better, improve digestion, normalize appetite somewhat (eliminate “munchies” from linoleic acid triggering the endocannabinoid system), and other little things. If you’re 300lbs and you only lose 1-2lbs in these 30 days, it’s time to add an intervention. You should be dropping fat like crazy. If you’re only 20lbs overweight and you lost 1-2lbs, you’re probably fine and don’t need to change much.
If you’re very overweight and aren’t dropping much fat just by cutting out PUFAs, pick either keto or carbo. I wouldn’t overthink it, just imagine which one you prefer intuitively: a heavy cream & beef diet, or a potato/rice diet with fruit? If one of these makes your mouth water, pick it. If they both sound disgusting, flip a coin. Try your chosen intervention for 30 days.
If you absolutely hated keto/carbo or it doesn’t agree with your digestion, try the other one for 30 days. If it was quite tolerable but you didn’t lose any more fat, you might need to try the other one or add another intervention.
There isn’t a big downside to unsuccessfully trying the other side of keto/carbo. Worst case is you lose 30 days and gain a few pounds.Add protein restriction to your keto/carbo diet for 30 days. On a keto diet, this means upping the (beef/dairy) fat and cutting down on meat. The easiest way, in my opinion, is to just eat lots of heavy cream. Add it to your coffee. Whip it up for dinner. Personally, I eat about 150g of fatty beef per day, and nearly everything else I eat is heavy cream.
If you’re adding protein restriction to “carbo,” you’ll have to substitute some fruit or similar sugar source for starch. Most starches contain about 7-10% protein, which is closer to “moderate” than “protein restriction” - especially if you then add any other protein on top. Most fruit & sugar sources (like juices, honey) contain almost 0% protein. The Kempner rice diet seems to have been about equal amounts fruit/fruit juice and rice, allowing Kempner’s patients to drop from the 7-8% protein in rice to about 4% protein overall.
How low should you take your protein? I think that <10% is probably necessary for any therapeutic effect of protein restriction. Going much lower becomes quite difficult, and besides Kempner’s 4%, my own 6% are the lowest I’ve seen in pretty much anyone out there. I’d say try 10%, and if that doesn’t work for 30 days, maybe give 6% a shot.
My prediction is that going through this “gauntlet” of 30-day experiments will help about 85% of people lose significant amounts of fat relatively effortlessly, eating to satiety.
The Potato Diet led the average participant to lose 10.6lbs in a month.
In my own ex150 trial, the 7 out of 10 participants who finished lost 9.6lbs in a month on average.
Now I do suspect that some people are even tougher “weight loss nuts” to crack and we might have to change or amend this dietary gauntlet in the future.
But I’m quite confident that it would help a lot of people reverse their obesity as is.
excellent summary of your work so far. maybe a Twitter-article version for easy sharing there, too, would help educate people.
If seed oils were invented in 1860, then your statement "Almost nobody restricted food intake 100 years ago [1925], and yet, somehow, almost nobody was obese." doesn't quite make sense. Seed oils had been around in increasing quantity for 65 years at that point. Crisco came onto the market in 1911. From their website: "In 1911, Crisco – the first-ever shortening made entirely of vegetable oil – was born. Clad in pristine white-paper overwrap, Crisco was seen as a more “pure” and economical alternative to animal fat and butter. Soon after our first print and radio ads debuted, products flew off of the shelves."
I do think seed oils play a part but the obesity epidemic wasn't really a thing until the 1970's. It seems more like a perfect storm including the usual suspects: pesticides, artificial colors and flavors, corn syrup, certain medical interventions, fat fear promulgated since the early 1960's and sun fear around the same time. I do think that getting seed oils out of one's diet is one of the most important first steps to take though- just doing that eliminates most of the other processed junk too.