I recently came out against paleo in the open thread, and realized that I probably haven't yet heard the strongest arguments in favor of a paleo diet. So, what are said arguments?
EDIT: Or more generally, why should I eat less carbohydrates and more protein / fat?
I'm very skeptical of reasoning like "it was like that in ancestral environment so it must be good". There are at least three reasons that makes me uncomfortable with the reasoning :
Even if we consider evolution to be a perfect optimizer (which it is not), there is a huge difference between "our digestion system is optimized to make the best possible use of food X" and "food X is best possible food for our digestion system". If you made an algorithm A optimized to transmit data on a noisy channel N, it doesn't mean the algorithm wouldn't run better on a less noisy channel C. There may be an algorithm B that work better on the clear channel C than A, but still, A can work better on C than on N.
Evolution doesn't optimize for the same purpose we do. Evolution doesn't optimize for us to live long, it has a very low pressure to make us live past ~60, for example.
We have completely different lifestyles and activities than we did during paleolithic. And the optimal diet very likely depends of lifestyle and activities.
That said, what would convince me to do a diet is not a plausible-sounding reasoning, but some evidence of short-term and long-term effects on a sane sample size, with a control group. Something which seems very rare in the diet field, saddly.
Your grounds for skepticism match the heuristic that Anders Sandberg and Nick Bostrom propose in The wisdom of nature quite closely. They propose this heuristic to evaluate interventions to enhance humans, but it's clear that it has much broader applicability. Here's the relevant excerpt:
Suppose that we liken evolution to a surpassingly great engineer. (The limitations of this metaphor are part of what makes it useful for our purposes.) Using this metaphor, the EOC can be expressed as the question, ‘‘How could we realistically hope to improve on evolution’s work?’’ We propose that there are three main categories of possible answers, which can be summarized as follows:
• Changed tradeoffs. Evolution ‘‘designed’’ the system for operation in one type of environment, but now we wish to deploy it in a very different type of environment. It is not surprising, then, that we might be able to modify the system better to meet the demands imposed on it by the new environment. Making such modifications need not require engineering skills on a par with those of evolution: consider that it is much harder to design and build a car from scratch than it is to fit an existing car with a new set of wheels or make some other tweaks to improve functioning in some particular setting, such as icy roads. Similarly, the human organism, whilst initially ‘‘designed’’ for operation as a hunter-gatherer on the African savannah, must now function in the modern world. We may well be capable of making some enhancing tweaks and adjustments to the new environment even though our engineering talent does not remotely approach that of evolution.
• Value discordance There is a discrepancy between the standards by which evolution measured the quality of her work, and the standards that we wish to apply. Even if evolution had managed to build the finest reproduction-and-survival machine imaginable, we may still have reason to change it because what we value is not primarily to be maximally effective inclusive-fitness optimizers. This discordance in objectives is an important source of answers to the EOC. It is not surprising that we can modify a system better to meet our goals, if these goals differ substantially from the ones that (metaphorically might be seen as having) guided evolution in designing the system the way she did. Again, this explanation does not presuppose that our engineering talent exceeds evolution’s. Compare the case to that of a mediocre technician, who would never be able to design a car, let alone a good one; but who may well be capable of converting the latest BMW model into a crude rain-collecting device, thereby enhancing the system’s functionality as a water collecting device.
• Evolutionary restrictions. We have access to various tools, materials, and techniques that were unavailable to evolution. Even if our engineering talent is far inferior to evolution’s, we may nevertheless be able to achieve certain things that stumped evolution, thanks to these novel aids. We should be cautious in invoking this explanation, for evolution often managed to achieve with primitive means what we are unable to do with state-of-the-art technology. But in some cases one can show that it is practically impossible to create a certain feature without some particular tool—no matter how ingenious the engineer—while the same feature can be achieved by any dimwit given access to the right tool. In these special cases we might be able to overcome evolutionary restrictions.
You should start the excerpt earlier to explain what is meant by EOC:
The starting point of the heuristic is to pose the evolutionary optimality challenge:
(EOC) If the proposed intervention would result in an enhancement, why have we not already evolved to be that way?
Gwern discusses these on his drug heuristics page.
I started recording my weight daily simultaneous with beginning the diet [...]
When I started my diet I went to the gym 4-6 days a week, alternating between running and weight-lifting. I currently go to the gym 4-6 days a week, alternating between running and weight-lifting.
So, you started going to the gym a lot and you started measuring your weight daily and and you lost weight and got healthier. (And also happened to change your diet.)
That's not really an experiment with Paleo. Not unless you'd already been going to the gym like that and paying that much attention to what you ate and how much you weighed.
Edit: I suppose that the "when I started" statement could be read two ways, one of which would imply that you already worked out that hard (6 days per week) prior to paleo. Though it seems odd you'd be able to do so and still be able to lose 20% of your bodyweight in fat, so I'll assume for now that's not what it was.
So, you started going to the gym a lot and you started measuring your weight daily and and you lost weight and got healthier. (And also happened to change your diet.)
You aren't interpreting that sentence correctly. Before and after starting the diet, I was exercising approximately the same amount. Although it is possible that merely increasing the frequency of recording my weight had a (round-about) causal effect on my weight-loss, the magnitude of the effect doesn't seem very likely.
Although it is possible that merely increasing the frequency of recording my weight had a (round-about) causal effect on my weight-loss, the magnitude of the effect doesn't seem very likely.
Measuring weight does lead to weight loss- but mostly through unconscious changes in diet and activity level. If you're consciously changing your diet at the same time, it should wipe out those effects.
People tend to like food that is harmful. And they tend to dislike diets that are potentially useful: most people fail to diet through failure of will, i.e. because the diet is not fun. So the evolutionary heuristic seems not to hold for food.
There are different theories why people like harmful food, but all of them try to explain the fact that they do like it. E.g., it is a purposely designed superstimulus due to competition between food sellers. Or, people evolved to want energy rich food that was rare in the ancestral environment, and when it is plentiful they actually overeat. Etc.
I think part of the problem is that the question being asked varies. It matters what your aims are, (weight loss, general health, etc.) and what you're comparing it to. People tend to conflate:
For aim X, is Paleo better than the generic unstructured diet I've had up to this point? Probably yes.
For aim X, Is paleo better than a calorie and nutrient controlled diet based on all the available evidence? Probably not.
And what you mean by supporting it.
Is the paleo 'philosophy' that you should eat like in the ancestral environment always going to be best? No.
Is "eat more meat and vegetables and less refinied carbohydrates" a useful guideline for most people? Yes.
When I ask this question I am usually referred to Gary Taubes. Also see Wikipedia. I don't have time to evaluate the evidence, but I'm pretty skeptical of nutrition science in general.
Edit: Eliezer has spent a few years doing the only thing you can do: try a bunch of diets on yourself and measure the results.
Obviously, the reason I tried and am trying multiple diets is that the experimental result is always that nothing actually works. Except that the Shangri-La diet worked for twenty pounds and then mysteriously stopped doing anything (i.e., abrupt cessation of the diet did not result in any significant change in weight trends) and Seth Roberts couldn't get it working again. Paleo was among the diets tried, and it didn't result in any weight loss or other detectable differences. Non-US-approved, powerful, dangerous drugs like clenbuterol, which are supposed to cause weight loss on the order of a pound a day, produced standard side effects but no weight loss in me.
I figure that metabolisms vary at least 10% as much as minds, which is a HUGE amount of variance. It actually points up something I may post about at some point, which is that statistical science itself is often a dead end - you can publish paper after paper after paper about effects that show up in 60% of the population - but you don't know what separates the 60% from the 40% - and still have no real grasp on the phenomenon and no real ability to manipulate it.
It actually points up something I may post about at some point, which is that statistical science itself is often a dead end - you can publish paper after paper after paper about effects that show up in 60% of the population - but you don't know what separates the 60% from the 40% - and still have no real grasp on the phenomenon and no real ability to manipulate it.
I'd be very interested in such a post.
It actually points up something I may post about at some point, which is that statistical science itself is often a dead end - you can publish paper after paper after paper about effects that show up in 60% of the population - but you don't know what separates the 60% from the 40% - and still have no real grasp on the phenomenon and no real ability to manipulate it.
Have you posted something about this? I think it's an important point, too little appreciated on LessWrong.
ETA: I see that I asked the same at the time. I guess that means the answer is "no".
Michael Rae--Aubrey de Grey's co-author and the most knowledgeable person about matters of nutrition and supplementation that I know--doesn't have a high pinion of Gary Taubes. Most of his remarks are spread over several messages on the Calorie Restriction Society mailing list; here's a representative sample:
More Calories cause people to gain weight and develop metabolic disruptions. Nothing Taubes presents meaningfully challenges this. And even if loading up on lard really didn't cause one to gain weight and doing the same with green veg (or even Wonder Bread) did, that wouldn't make loading up on lard healthy.
Taubes is very good at picking apart the weaknesses in the epidemiological case against saturated fat (tho' the recent meta-analyses, particularly in the breakdowns by poly:sat ratio, significantly beat back much of the case there); what he hasn't done is either make a good case to the contrary, or even propose a credible alternative means of establishing the truth. You can only rely on the evidence you have, not the fifty-year, thousand-subject, randomized metabolic-ward diet-heart clinical trial on which Taubes wants to insist before he'll drop the pork rinds.
Well, in my case I've found that The Hacker's Diet pretty much works for me: the number of calories I get seems to be way more important than where I get them from (provided I get them from something reasonable, as opposed to 2000 kcal's worth of Coke per day or something like that).
It looks as though some people are healthier if they eat some grains. One of my friends has even found that his digestion doesn't work well unless he eats a good bit of wheat. He hasn't experimented with other grains that include gluten. He's got a couple of relatives who show the same pattern.
I believe it's important to do your own experiments.
A search through the NIH site does not seem to show a single reliable study on the subject...
The most obvious, and significant, argument in favor of paleo is that it tends to make it more difficult to overeat or maintain a lot of weight (i.e., it's acts as a fail-safe against akrasia). Imagine if you only had nuts, fruits, and vegetables to snack on and for anything else you had to cook a meal.
A better diet would be to push your eating window to 2-3pm to 10pm (intermittent fasting), and only eat low calorie things which you've cooked (and make sure you don't buy anything that can be eaten without being cooked).
Basically, from the research I've read, akrasia is the most significant factor in maintaining a healthy weight; anything you can do to force or impel your future self to commit to your non-akrasiac desires is a good idea.
I switched from vegetarian to paleo based on Food and Western Disease, which had the best recommendations I could find at the time. I will take Dr. Linderberg's recommendations as what "paleo" means. Oversimplifying:
Yay: Lean meat, fish, veggies, root vegetables, fruit, some nuts. Drink water.
Boo: Grains, dairy, refined fats, sugar, beans.
The primary argument in favor of the 'yay' foods is nutrient density: you get a lot of micronutrients (and protein) per calorie. For instance, you want to consume about 100mg of calcium per MJ of energy you consume. Spinach has 1517mg calcium per MJ of energy. In general, vegetable have lots of micronutrients and not very many calories, so they have lots and lots of micronutrients per calorie.
There are two main arguments levied against 'boo' foods. First is that they tend to have low nutrient density. Bread (whole grain) only has 60mg of calcium per MJ of energy. Second, they contain antinutrients, ie. chemicals that bind to micronutrients so we can't absorb them. Phytic acid is an example of an antinutrient. Nuts also contain phytic acid, which is why you'll notice a recommendation against eating a lot of them.
The second approach is to copy examine.com's answer, which deals specifically with weight loss, as opposed to general healthy eating.
> Many diets, both fad and more long-term diets, do work. This is mainly because they reduce calories. > When people switch to a paleo-lithic (hunter-gatherer diet), the foods they switch to are naturally more filling (higher protein, fiber, water content) or have less calories for the size of the food eaten (due to water content; a pound of broccoli has less calories than a pound of grains). > When people switch to a ketosis diet (very low carb), the higher fat and protein levels naturally provide satiety and fill people up. Also, there is some evidence that obese people have a maladapted response to serotonin (of which carbohydrates aid in the synthesis of) and thus omitting carbohydrates omits cravings. People lose weight on a ketosis diet because they eat less on a day to day basis, and avoid large binges caused from carbohydrate cravings. > Diets that manipulate fasting (Intermittent Fasting, Alternate Day Fasting) may have some benefits on the 'calories out' side of things as prolonged fasting might increase heat expenditure, but the most significant means for weight loss here is that you control eating. It is much harder to overeat in 8 hours than it is in 16.
These are the arguments that have convinced me--first to do paleo, and then to completely abandon it and eat soylent. There's also evidence that when populations switched from ancestral to western diets, their health declined. I think this is interesting evidence, and gives us a reason to test ancestral diets. I'm not aware of enough sufficiently good testing to make particularly good arguments for any diet.
I eat paleo whenever I can't have soylent. This means "whatever Lindeberg said," (outlined above). This is not antagonistic to carbohydrates. In fact, Lindeberg's best know for studying the Kitavans, an ancestral population that gets 69% of its calories from carbs. If paleo means "whatever Lindeberg said," then I've presented the strongest arguments from his book. This set of foods tends to be high in nutrients, low in energy, and devoid of antinutrients, so we get the nutrition we want in not too many calories.
However, if "paleo" means "replace carbohydrates with protein/fat," I can't help you. I no longer frequent the paleosphere and don't care which definitions you map to which diets. I've presented you with the strongest arguments I know for the diet I associate with "paleo". Sorry if that's not the diet you were looking for.
If paleo/primal advocates were strong rationalists, they would state the basic premise of the diet as follows:
"In the absence of evidence, the safest bet is to go with an ancestral diet."
And I think it is difficult to argue with that. When the evidence is mixed on a topic, it seems natural to default to tradition. In general I think paleo/primal dieters consider non-ancestral foods guilty until proven innocent...as in, you would have to present them evidence that a non-ancestral food was harmless and healthy before they started eating it. Most people are asking "Why shouldn't I eat X", but within the paleo/primal philosophy, the burden of proof rests on the advocates of novel foods, since they are the one who deviate from the ancestral default. If you want to debate what exactly the ancestral default was, that is an entirely different can of worms...but it is easy to say which foods are certainly not ancestral.
By the way, carbohydrates are not criticized by the paleo movement. You're thinking of Atkins. Fruits and tubers were part of our ancestral diet, and they are rich carbohydrate sources. They are encouraged in moderation.
The paleo criticism is that the modern diet's primary source of calories is now grain, which was scarce in the ancestral environment. More dedicated paleo folks will also attack milk, legumes, and nightshade vegetables (in that order) but really the majority of complaints are directed at grain. Before agriculture, no one really ever ate it. After agriculture, the population boomed and people didn't really have any better options other than grain. Now grain has been ingrained into our culture (pun intended) even though we have the resources to avoid it.
Anyway, there are three compelling pieces of evidence against grain: Lectin, Gluten, and Phytic Acid
1) Grain contains large amounts of phytic acid. Nuts contain more than wheat, but we don't typically eat that many nuts per day. Chemists would call it a chelator, something that binds with ions. Dietitians would call it an anti-nutrient...it essentially stops your body from absorbing Zn, Fe, Ca, and Mg. Ruminants and grain-eating animals have evolved mechanisms to break it down, but humans have not.
2) Lectin is a natural pesticide which irritates gut linings. Over time, it causes leptin resistance...which contributes to unhealthy fat storage patterns and other metabolic disorders like diabetes.
3) While the gluten free craze is a bit out of hand, there is increasing evidence for a gluten sensitivity spectrum. Not everyone is a celiac, but a substantial proportion of the population experiences some amount of gut inflammation via gluten.
Add to these three things the notion that grains make it easy to consume large numbers of calories and can cause sudden spikes in insulin, and I think you have a pretty good idea of why the paleo/primal diet discourages extensive grain consumption. It's really more about toxin and anti-nutrient avoidance than it is about carbohydrate vs fat as a calorie source - although it is generally agreed in these circles that modern diets are a bit carb heavy.
Hunter-gatherers eating traditional diets had very low rates of the modern "diseases of civilization" including cancer, heart disease and diabetes. When, however, hunter-gatherers switch to eating modern diets they start getting these diseases at high rates. (Of course this correlation doesn't prove causation, but still...) Also, anthropological evidence from bones show that populations usually became less healthy after adopting agriculture.
Although hunter-gatherers had lower life expediencies than we do, this was mostly due to them dying of stuff (murder, infections, viruses) that doesn't significantly reduce the life expectancy of Americans. The paleo lifestyle tries to combine the health virtues of modern day and paleolithic man.
I look at paleo as providing Bayesian priors for the health value of different foods. If food X was eaten for a million years by my ancestors whereas food Y was eaten for less than 20,000 years then my prior belief is that food X is probably healthier than food Y. This seems to be the implicit approach taken in the Perfect Health Diet. The value of paleo priors is inversely proportional to the strength of nutrition science.
Forgive me for not having citations to backup these arguments. They are based on my readings of the secondary paleo literature.
I have a related diet question on portion size; I guess I'll ask here rather than the open thread or "stupid questions" thread. Is spreading out your carb intake over the course of the day as good as eating fewer carbs?
If the problem comes from glycemic load (spikes in blood sugar & insulin levels), then it seems like it should be. Instead of eating a big slice of pie in the afternoon, I could have half of the slice then and half at night. Or, even better, a bite of pie every 20 minutes. Then my body wouldn't be overwhelmed by a huge burst of pie all at once; the constant snacking would mimic the slow-release pattern of low glycemic index foods.
We may soon be getting some real data. From Gary Taubes' blog: http://garytaubes.com/2012/01/updates-for-2012/
"Among the projects we have in the works is a non-profit, the Nutrition Science Initiative (NuSI), to raise money for the kind of research we think is necessary to clarify the relationship between dietary nutrients, obesity. diabetes and their related chronic diseases. We have a specific plan of research to pursue (or rather to fund so that established, unbiased researchers can then do the studies) and have already recruited a world class scientific advisory board and executive board."
According to a later post: "We’ll say much more about this when we formally and publicly launch NuSI in early September. The ultimate goal is to create what would ideally become a kind of Manhattan Project of Nutrition: a concerted, directed, well-funded research effort composed of the best scientists in the field — all independent and suitably skeptical — working together to generate the evidence necessary to put to rest, one way or the other, all the major and many of the minor controversies in nutrition research."
I'm very skeptical of reasoning like "it was like that in ancestral environment so it must be good". There are at least three reasons that makes me uncomfortable with the reasoning :
Even if we consider evolution to be a perfect optimizer (which it is not), there is a huge difference between "our digestion system is optimized to make the best possible use of food X" and "food X is best possible food for our digestion system". If you made an algorithm A optimized to transmit data on a noisy channel N, it doesn't mean the algorithm wouldn't run better on a less noisy channel C. There may be an algorithm B that work better on the clear channel C than A, but still, A can work better on C than on N.
Evolution doesn't optimize for the same purpose we do. Evolution doesn't optimize for us to live long, it has a very low pressure to make us live past ~60, for example.
We have completely different lifestyles and activities than we did during paleolithic. And the optimal diet very likely depends of lifestyle and activities.
That said, what would convince me to do a diet is not a plausible-sounding reasoning, but some evidence of short-term and long-term effects on a sane sample size, with a control group. Something which seems very rare in the diet field, saddly.
Your grounds for skepticism match the heuristic that Anders Sandberg and Nick Bostrom propose in The wisdom of nature quite closely. They propose this heuristic to evaluate interventions to enhance humans, but it's clear that it has much broader applicability. Here's the relevant excerpt: