The best kittens, technology, and video games blog in the world.

Thursday, March 26, 2015

Rationalist reasons why you should eat (some kind of) paleo

It could be plausibly argued that this is (some kind of) paleo

Let me present a really solid rationalist case for "(some kind of) paleo" diet.

We know shit about what people should eat

There are zero serious RCTs on long term effects of various diets in humans.

Non-human studies are close to worthless, as humans have extremely unique diet in all the animal world, so the best they can find is if something is actively poisonous.

There are some short term RCTs, but that's not generally the problem with food - we have reasonably good idea what kind of diet is going to get you in a hospital in a few weeks. Most problems people have take years to develop, and apparently nobody can afford to seriously study that. They generally don't find any actual health outcomes, just data mine for something that might plausibly correlate with some risk factor.

There's a bunch of observational studies, but that's close to worthless in principle, even if they find difference in health outcomes, and usually they just do data mining for correlates.

The original "Seven Countries Study" which started modern diet "science" can't seriously be described in any ways other than fraudulent, and little changed afterwards.

Scientifically, we know shit.

Experts know shit about food

As result of previous point, the so called "experts" know as much as 18th century bloodletting doctors knew about medicine, and following advice they pull out of their asses is more likely to cause harm than do good.

For a random example - you remember how the "experts" declared saturated animal fats evil, and they got replaced by "healthy" partially hydrogenated vegetable oil? Yeah, it's the same people.

This isn't really unusual that entire discipline of science is harmful, even if they know more than zero. When do you think actual doctors started doing more good than harm? Various dates between Pasteur's germ theory and invention of penicillin are plausible, but it's clear as hell that for thousands of years doctors were actively harmful, and yet people still treated them as "experts" all that time. Seriously, people like this were considered "experts", just as today people treat former trans fats pushers as "experts". And it's not that doctors back then knew nothing - they had a lot of information about human bodies, it's just that they extrapolated from it in a way that ended up ridiculously harmful.

We will probably never know what we should eat with any kind of scientific rigor

Even if we suddenly started doing high quality long term RCTs (and nothing indicates that we will), they are really hard to do. Even if we actually did RCTs back in 1950s and figured out for sure what people in 1950s should have eaten to be healthy today, by now food available changed so drastically that this information is probably not very useful.

Even if food is labelled the same as before - like we have something we call "chicken" and they had something they called "chicken" - its nutritional value might have changed drastically, so even solid evidence would get less and less applicable with every year unless you insisted on avoiding all recent food (I'll come back to that thought).

You still have to eat something

Hopefully you're with me so far - we don't really know what kind of food is good for you long term, at best we know what kind is bad for you short term (nothing that we know of has any major positive short term effects).

This is the point where the rational thing would be to say "I don't know", and move on. Usually, that's sounds strategy - you might not know what was Julius Ceasar's favourite color, what dark matter is made of, or if Riemann hypothesis is true, and fortunately you don't have to know everything.

No such luck here. Even if you don't have solid answer you still need to eat something, and unless you're completely unconcerned about your health that means committing to a hypothesis you find more plausible than alternatives, even if it's very far from epistemic standard you wished for.

You should not eat what everybody else is eating

The most important technique of rationality is the Outside View. It basically means if you do what everybody else is doing, you can expect similar results to results everybody else is getting.

So if you're going to eat just what everybody else is eating - you're fucked. And I'm not just talking about being a bit heavier than you'd like - you're likely to be really fucked. 40% of Americans age 60 or older have diabetes already or are prediabetic, and you can safely bet which way this number is going. Epicenter of this might be in America, but it's spreading to the rest of the world fast.

Now, admittedly, that's a fairly old age, and with proper drugs diabetes and other metabolic fuckups can be managed, so if you're fine with that, you're welcome to eat like everybody else. In case you aren't, keep reading.

Oh, and if you think you should follow "expert" advice anyway, even if it's mostly made up, that is actually what most people have been doing. They indeed have been trying to shift their diets in directions advised by "experts" (admittedly not all the way, just gradually), with disastrous consequences.

Alternative dietary advice is mostly bullshit

So what should you eat instead? Mainstream dietary advice is bullshit with very little scientific basis, and sadly the same can be said about most alternatives - those that even try that is, a lot don't even bother, and just go for New Age bullshit and anecdotal evidence.

There is one exception, sort of

There's some hope. Metabolic fuckups are something modern humans suffer from - but until recently they were very rare, and they were not seen in animals in the wild either. And by the way modern metabolic fuckups start affecting animals living with or near humans too, so don't even think about blaming humans specifically for being too lazy or weak-willed or whichever bullshit people blaming the victim come up with.

People managing to avoid them for so long without actively trying is reasonably good evidence that they were doing something right. Now sadly they didn't follow proper research protocols, so our extrapolations of what they were doing right are not as solid as we'd like, but it's a lot better starting point than alternatives.

Is this endorsement of full paleo?

Not really. You can't really eat paleo, and it's not clear you should even try if you could.

First, there's epistemic issue. We don't really know what people ate back in Paleolithic - well, we know some things they definitely didn't eat, but from information we have it's clear their diets varied drastically from place to place, and they were all fine anyway (at least as modern metabolic fuckups go).

Second, pretty much none of the foods they ate back then are available. Modern chicken or pork meat not particularly close to wild meat (assuming you'd be willing to eat all organs as well for better match), farmed fruit and vegetables are drastically larger and less nutrition dense than ones found in the wild, and even modern wild-caught fish are a lot more exposed to methylmercury and who knows what else we've been dumping into the oceans than Paleolithic fish.

Third, nobody even seriously tries. Humans evolved in Africa, and didn't move out futher than the Middle East until fairly recently, and I haven't seen anybody who takes paleo so far that they only eat African (and maybe Middle Eastern) food. And New World food? Humans have been eating grains longer than tomatoes, pineapples, or turkeys. Nobody seems to care.

You can't eat real paleo, the best is some vague approximation.

Not that it matters, even people who claim to eat paleo, can't seem to agree what counts and doesn't count as "paleo". Seriously, go to any "paleo" forum, you'll see ridiculous claims on what is or isn't "paleo".

Why go all the way to Paleolithic?

On the other hand, you don't need to go all the way to Paleolithic. People have been eating farmed food for thousands of years, and for all the problems they had, modern metabolic fuckups were not among them.

There's less evidence farming-era food is healthy compared with Paleolithic food. Especially earliest generations of farmers are be rather unhealthy, but that could plausibly be blamed on adjustment issues with every new technology. If you took two healthy groups of Paleolithic hunter-gatherers and swapped their places, both would suffer for generations until they figured out how to thrive in their new environment. Early farmers had very limited choice of food, and it took civilization very long time to figure out how to safely live in more crowded conditions.

As far as we know early farmers from early pre-industrial times were doing reasonably well - and then industrialization fucked up everybody's health due to bad sanitation in cities, which is by the way another example of adjustment issue that got completely resolved a few generations later.

The upside of extending your dietary choices to traditional farming-era foods is that you can actually get such foods, or reasonable approximation of them. Strict "paleo" diets are really playing pretend with food.

How strict should you go?

Even with modern junk food, a bit less than half of people end up with fucked up metabolism. Most end up no worse than slightly overweight, which is more aesthetic than medical problem, at least as long as it doesn't go any further.

If you eat mostly traditional food, and only occasionally something modern and dubious, you're probably going to be just fine. Eating healthy food is not going to turn you into a superman, it's just how you avoid certain failure mode. Personally, I'd still avoid anything with any amount of partially hydrogenated vegetable oils, or other obvious junk ever, but even that probably won't screw you unless you make that a habit.

If sticking to pure paleo or whatever helps you resist temptations better (assuming you even find modern food tempting, most isn't particularly tasty and offers mostly low price and convenience), there's nothing wrong about it.

What if it's wrong?

It's possible this is all wrong. It would be ridiculously irrational to pretend certainty, and I'm definitely not doing that. Maybe it's not really food, but antibiotics overuse, sedentary lifestyle, exposure to lolcats, God's punishment for gay marriage, side effect of government's mind-control rays, or who knows what else.

Or maybe it's food, but it's some aspect of it that trying to eat "some kind of paleo" doesn't fix. It's not like you know for sure what they've been feeding those organic chickens, and maybe that's the thing that's going to screw your metabolism.

In the end, it's a bet. Fortunately the burden of eating "some kind of paleo" is not very high, and it has a decent chance of working. That's not something that can be said about the alternatives which are either very burdensome, or unlikely to help you, or both.

No comments: