This blog is no longer active, you can find my new stuff at:,,, and

The "It’s really complicated and sad theory of obesity"

Epistemic status: Heavily speculative rant

Reader note: Please keep in mind I'm talking about obesity, not chubbiness. Chubby people (slightly overweight BMI or mid-range overweight BMI but muscular), provided they don't suffer from metabolic syndrome, are perfectly healthy, potentially more so than their slim counterparts.

i - Muscle

There are more muscular people walking around nowadays. Not sure about the mean, but the upper extreme is way more extreme. Schwarzenegger was the known limits of the human physique in the 80s. Now you can walk into any half-respectable gym and find someone that buff. Modern outliers in bodybuilding look outright inhuman, tiny people walking around with <3% body fat and muscles so large it's hard to tell how they can function

Why is this the case? What caused this?

There's probably a variety of factors. For one, there might be changes in the kind of food available and nutritional [bro]science. There are more supplements and they are better. Steroids are stronger, safer and more socially acceptable. There are more gyms and more of them are focused on hypertrophy and strength training. There’s the social aspect, Schwarzenegger was viewed as a bit of a weirdo, nowadays there’s a lot of entourages that respect and even celebrate people aiming for that level of physique. Not to mention that people seem more attracted to muscles now, even in the extremes, in both men and women.

Then we have cultural changes, muscles are more fetishized by media, kids are encouraged to go to the gym, movie stars are often required to look a bit buff; Heck, it seems that even famous writers, politicians, businessmen and scientists are doing it. Last but not least, modern people “know” how to get to a buff physique; Back in the day you were stumbling in the dark, now there are coaches that have walked the path, that came up with methods of achieving this, models that you can follow.

You could try to find THE THING that leads to people being more buff, and presumably, you could draw a lot of correlations, very strong ones.

Look, the muscle mass of Mr Olympia proportional to height is almost perfectly correlated with the number of tensor cores in a GPU / with the increase in average protein consumption / with the number of cities that add fluoride to drinking water.

It’s obviously vitamin D, no doubt about it, there’s almost no bodybuilder that’s deficient in it, and most of the best ones have very high vitamin D levels.

Gym density is what does it, the correlation between the number of bodybuilders and gyms in a city is undeniable. Both across cities and longitudinally within a city. The trend even holds fairly well across countries.

Or you could try to be “scientific” about it and run studies of 1, 4 or even 20 weeks trying to get people to gain more muscles, then come up with conclusions like:

Hyperbaric oxygen is the best way possible to gain muscle.

Leucine supplements are correlated with more muscle definition over 1 to 14 days of consumption, the effects last 2 months into the study.

Doing clean jerks is scientifically proven in a randomized controlled trial with 5 other types of lift to boost biceps size by 1.4-1.8 times as much.

But even the average brodude would laugh at all such findings. They are not "wrong", but they are hardly the key to hypertrophy; Gaining muscles is hard, you can use various tricks, but they level off over time, and they work differently in different populations with different levels of training. Everybody knows a “newbie” in their early 20s can do anything and gain muscle, while someone experienced in their 30s will be lucky to gain 1kg a year if they are going natural and pushing hard.

Professional bodybuilding, and even maintaining physique as an amateur, is not a one-variable problem, it's a lifestyle, it's hundreds of factors intermingling, it's something that’s personalized, that require you to experiment. You have to figure out the things that work for you.

ii - Complexity

The same can be said about any other complex trait of the human body, be it the length of hair, the thickness of beard, bone density, running speed, ability to solve math problems or respiratory capacity.

Even more importantly, almost every single trait gets less mutable as we age. If you haven’t put any muscle until age 40 you might be able to gain a bit, but you’re never going to be anywhere close to someone that started in their 30s, who’s not going to be anywhere close to someone that started in their 20s. We get accustomed to lifestyles that lead to certain traits and changing those is very difficult, both because it requires changing habits, and because of the innate mechanisms that adapt to maintain those features.

The amount of fat one has is as complex a trait as the amount of muscle. Yet I see people all around trying to come up with “theories” about obesity. It’s the mercury in the fish, the sugar in the bread, the lithium in the groundwater, the α-linoleic acid in creamy spreads, the lack of jogging, the CO2 in the air.

People try to study weight loss over a few days or weeks and have the audacity to think their findings will generalize to a year, or even a lifetime.

This seems insane and unscientific, if you search hard enough you’ll find a correlation and complementary theory for literally anything.

Yet, pun not intended, people eat this stuff up.


Obesity is a very complicated thing. The physiology of obese people is a lot different from that of normal-body-fat individuals. From their androgen hormones to their insulin, leptin and glucagon response, to their fat cell genotype, phenotype, fatty acid composition, to their caloric intake, both overall and tissue-specific, to liver function, small intestine and stomach size, gut flora, mouth flora, bone composition, brain networks, and the expression of thousands of proteins in virtually any tissue you chose to study.

There’s also a huge lifestyle difference between obese and normal-body-fat people, from exercise patterns to sunlight exposure, to hobbies, friend groups and cultural preferences.

Does being a really complex trait guarantee there's no "easy" cause or solution to it? Not really.

Dentine and enamel growth are complex things, but you can just add fluoride to water and toothpaste for incredible results in terms of oral health.

Both IQ and violent tendencies are really complex traits, but you can remove lead pollution and the former increases while the latter decreases with astounding replicability.

Getting lung cancer is a very complicated process but you can still eliminate almost all cases with smoking cessation.

Will obesity turn out to be part of this category? I don't know, prima-facia it seems like the closest comparison would be with something like lead poisoning. Where removing one or two things might have a huge impact. Given how distributed obesity is among populations in widely different environments I doubt this.

But, as with the previous examples, even if a very useful intervention is found, there's little reason to think it will be able to revert most damage for people that are already obese.

iii - Sliming

So getting slim might be best viewed as something akin to building muscles:

This is not a popular theory of obesity because it kinda sucks. I can’t sell the “it’s impossible, give up" diet book and complimentary supplement subscription. I can’t publish a paper saying “Hey guys, turns out it’s really complicated and nobody has any idea”. The lack of incentives means the only “evidence” for this theory is common sense; It’s the fact that all other theories have failed, and that “amount of body fat” is such a complex property interwound in so many feedback loops, that I’d be nothing short of miraculous if you could easily modulate it by a lot.

For some reason, we’ve collectively failed to adopt this as the null hypothesis for obesity. I’m not sure why, but we’re currently stuck with a “CICO” model, where people can magically decide to just “eat less” without either going completely mad or just becoming bereft of energy and losing weight way too slowly to tolerate.

Even worst, there are a few outliers for which on-trick solutions work... fast, eat keto, count calories and eat 400kcal less each day, go vegan, go jogging, whatever. I suspect that a lot of these outliers are not the result of one trick, but rather just people omitting dozens of behavioural changes, or attributing those behavioural changes to their “one trick”. But, one way or another, these interventions don't scale well.

The implications of this are horrible, in so far as obesity and metabolic syndrome are really bad. Bad in that they lower your life expectancy considerably. Bad in that they add morbidities that make you enjoy your shorter life much less. Bad due to a variety of social reasons, which, while potentially unwarranted, won’t go away anytime soon.

If we accepted the “it’s really complicated” default hypothesis, we’d be taking obesity more seriously. To me, the reasonable thing to tell a man in this 30s that’s obese is something like:

Look, you need to start taking some appetite inhibiting drugs right now. You have to schedule liposuction, then a gastric bypass. You should start a temporary course of hormonal therapy and couple that with ~2-hour gym sessions 2 to 4 times a week to gain muscles to allow you to consume the extra calories you’re eating. Also, take these drugs that might help because . Vastly alter your diet, cut literally all added sugar, and very sweet fruits, and most vegetable oils and most high-calorie good tasting food; Ideally switch to the most “conservative” definition of a good diet you can tolerate, something like salmon, rice and broccoli. And after a few years, with a bit of luck, you’ll have a normal weight. Or, at least, you’ll be just chubby, which is fine.

Note: The above is for illustrative purposes, do not try it. Were I to be obese I would not be doing the above, nor would I recommend it to a friend. If you want to do an extreme intervention to fix obesity please consult with the few doctors or books that work in this space.

The above sounds extremist and discouraging, it sounds somewhat hard and dangerous, and expensive, it’s a life-altering intervention in a lot of bad ways, not in the fun “let’s use almond flour and eat loads of butter” kind of way. However, it’s just a combination of most non-overlapping interventions that have some long-term effect on fat loss.

Even more importantly, this correct view might have the positive effect of ringing alarm bells for how important it is to lose weight early in life, or never gain it.

It seems like one reason people get obese is that they see it as temporary. You have people coping using food during stressful periods like PhDs or residencies or whatnot, and assuming that they will just be able to lose the 15 extra kgs; Even when all of reality is telling us that putting on weight is a vicious cycle, it becomes easier to gain and harder to lose. Having obese children seems like child abuse, or at least bad parenting equivalent to feeding a 5-year-old a diet of bourbon and cigarettes. But people don’t seem to care about this, because putting a child on a diet is hard, and they have this background idea of “Oh when they get older they can decide to lose weight”.

Again, this is not to say that finding causes for obesity that we can implement policy upon is good.

Deincentivinzing mono or poly saccharides would be good. Doing the same for linoleic acid would also be good. Reducing certain pollutants would be too. As would incentives to exercise. But an aggregate improvement does not an individual solution make.

And the reality, as it seems obvious to me, is that obesity is currently mostly incurable at an individual level, and not stating that fact plainly makes the problem worst.

Thanks to SMTM for their feedback on this article. Also shout-outs to Trevor Klee, since reading his blog has inspired articles like this one.

Published on: 2021-10-25



twitter logo
Share this article on twitter
 linkedin logo
Share this article on linkedin
Fb logo
Share this article on facebook