Diet Advice?

by Vaniver4 min read12th Jul 201139 comments


Personal Blog

So, I know a number of friends on Paleo who recommend it. I recently read through a lot of bulletproofexec, who recommends his own variant of paleo. I care about my health, and so I need to resolve my diet and their advice somehow. Summarized data points:

  1. My diet is about 80% wheat, and the rest is sweet potatoes, exotic grains (like amaranth or quinoa), rice, margarine, honey, cheese, and meat. The primary principle of paleo is that Grains Are Bad, and accepting that premise would require a radical restructuring of my diet. Right now I bake a loaf or two of (sourdough) bread a day, eat pasta about every other day, eat at least some raw sweet potato a day, eat cereal every other day, boil quinoa and rice every now and then, and eat meat only when I go out with friends.
  2. I strongly prefer wheat to rice, and rice to corn (also, white wheat to whole wheat). I enjoy the bread I bake enough to eat until I physically can't store more food in my stomach. When I bake bread with 75% wheat and 25% amaranth, it tastes noticeably worse (which I can tell because I eat what I cut and don't try to eat it all).
  3. I am rarely sick (maybe something on the level of a sore throat once a year), and I've been intelligent and lively my entire life, and have been on a meat-heavier version of this diet (less bread, more pasta, more chicken) until recently.
  4. To the best of my knowledge, I do not experience meat cravings. I chewed ice growing up (which suggests an iron deficiency), but this habit mostly stopped a few years ago. I have craved fat once or twice in the last year, which I responded to by whipping up and eating a batch of cookie dough. Most of my friends on paleo experience bread/grain cravings about once a week, though my friend on a slow carb diet (which allows him to eat bread once a week) doesn't.
  5. I'm 6'0" tall and weigh 160 lbs (182 cm / 73 kg) for a BMI of 21.7, which is the exact middle of the "normal" range. I have a sedentary lifestyle with no regular exercise besides walking (which has been reduced to about once a week thanks to summer), and thus am not as muscular as I would like. I used to weigh about 180 lbs, and a change in diet (from eating out frequently to eating in frequently, the replacement of meat with bread, and a general reduction in portion size) dropped that down to 160 lbs with no exercise. A brief experiment with creatine supplementation and high-intensity exercise ended well, and so I plan to resume that soon.
  6. I have a poor sense of smell and thus am not good at telling foods apart; I have a preference for simpler foods and can't tell a difference between chicken and steak. I'm also pretty xenophobic when it comes to food, but have gotten better at experimenting with new tastes.
  7. I get warm fuzzies from eating little meat (for resource conservation reasons) and from eating cheaply (my grocery bills come out to about $3 a day and I eat out 1-2 times a month), but I imagine those would be outweighed by small health / intelligence boosts.
  8. My lactase production disappeared about age 20, but I respond well to lactase pills.
  9. Bulletproofexec and one of my friends on paleo both have wheat allergies.

I find the logic behind paleo questionable. Yes, hunter-gatherers are adapted to a different diet, but fire was first used to cook food 2 million years ago, and appears widespread by 100 kiloyears (ky) ago, with noticeable adaptations in humans (from smaller teeth to resistance to air pollution). Lactose tolerance demonstrates the ability of human biology to adapt to new diets. Civilization dramatically speeds up evolution- it probably took about 25ky for European hunter-gatherers (and later farmers) to go from a mean IQ of 85 to 100, and about 1ky for urban European Jews to go from a mean IQ of 100 to 115. Am I really supposed to believe that there aren't genes floating around that wheat (domesticated 10ky ago) is good for?

My interpretation of this data is that my current diet works well for me, and paleo is unlikely to work better. I am willing to experiment, though- if I will actually live better on a different diet, there is little holding me back besides a lack of information. My values, in descending order of importance, are: brain function, overall health, appearance, mood, and cost. (Note that those are weights- something can improve brain function but be so costly in dollars, prep time, and terrible taste that I'm not interested.)

So my question for you is: Should I try paleo (more likely, the bulletproof diet)? If I do, what data should I collect? Better yet, what data can I collect now to determine if I have any nutritional deficiencies?


39 comments, sorted by Highlighting new comments since Today at 4:09 AM
New Comment


I'm a pseudo-paleo dieter since 2009. Outlining my diet considerations in a nutshell:

There is a tangled hierarchy of data when it comes to nutrition. On the top there are the large-scale longitudinal epidemological studies, which tell us about health effects directly; however, they are by far most prone to bias. Then come smaller clinic trials and placebo controlled studies, and after that are "dietary arguments", like paleo. Higher tier data can disprove lower tier data, but because analysis of higher tier data is often very biased (see e.g. Campbell's take on the China study), we have to reach back to more trustworthy lower tier data to filter out absurd results (e.g. if a small clinic study finds that substance X is toxic in specific ways in dose Y, then we're compelled to question a longitudinal study's result that prolonged consumption of X in dose ~Y correlates with longer lifespan.).

As to the paleo argument, a stronger version of it posits that a diet is optimal if it closely resembles the diet that was consumed for the longest evolutionary time.

I mostly reject this argument.

First, there are "new" foods that happen to be okay because their specific composition and nutritional profile contains nothing obviously harmful. Take butter: it is just a mixture of fatty acids which mixture doesn't fundamentally differ from actual animal fat. Take oat: it is grain that happens to be okay because its proteins are positively harmless compared to wheat's gluten and lectin, the former inducing inflammation and possibly leaky gut and the latter messing up one's fat metabolism (plus there are phytates that demineralize teeth and make minerals from food biologically unavailable). Arguments for irreducibility of diet (for example: "whole pieces of food and whole diets have effects irreducible to the effects of constituents") sometimes apply but more often does not; a specific fatty acid in butter or tallow is just the same thing, for instance. If a food has benign constituents that are known to interact in non-harmful ways, it really is a strong indicator that that food is benign.

Second, evolutionary pressure is not constant, and it was most likely much greater in the 10k years since agriculture than in the 10k years before it. This is of course in the same class of arguments as paleo itself; you cannot just consider yourself fully evolved to deal with whatever food; you have to also discount this argument with the prior information you have.

I accept though a Weak Paleo Heuristic: "closely scrutinize foods that are literally evolutionarily unprecedented in terms of types of constituent macronutrients or the ratios and amounts of those nutritients". This heuristic aims a lot more specifically at industrial, 20th century foods than the standard paleo argument. It picks out trans fat, vegetable oil and refined sugar, which are indeed rather bad. It ignores wheat though, and wheat's significant disadvantages are as of now quite clearly and forcefully established (see e.g. this).

I singled out the types of foods to my diet that are least likely to be harmful or lifespan-reducing. I kept eggs, high-fat dairy, most meats and fish, coconut oil, cocoa, oat, vegetables, most fruits except for the very very sugary ones, and kept some benign starchy food in moderated amounts, like lentils, peas and potatoes, and I occasionally eat rice. I minimized (eliminated when possible) wheat, simple sugar and n-6 fatty acids.. I'm also on a carefully constructed regimen of supplements, of which the most important are probably Vitamin D3 and fish oil, the others being only of interest to life extension aficionados.

My interpretation of this data is that my current diet works well for me

Do you have a comprehensive blood test? If you only have a subjective account of your well-being, that is extremely limited data. A lot of humans get by subjectively well on bad food until their late 40s, when they rapidly start deteriorating. A lot of sneaky deficiencies and disorders can be discovered with a simple blood test, and it also allows you to make a reasonably good estimate of your cardiovascular risks, although the interpretation of blood lipids is also prone to great amount of bullcrap as of now (e.g. last time I got a warning on my total cholesterol because my HDL was too high).

My personal subjective well-being did not radically change when I switched to my pseudo-paleo. My blood profile is hard data however: I have gone to a risk level lower than the 5th percentile. If you want to optimize well-being, you should still at least try some diet modifications. Wheat elimination and Vitamin D supplementation (2000-5000 IU per day on average, optimal amount is calibrated with blood testing) are the two single modifications that are most often reported having dramatic effects.

Yes, hunter-gatherers are adapted to a different diet, but fire was first used to cook food 2 million years ago, and appears widespread by 100 kiloyears (ky) ago, with noticeable adaptations in humans (from smaller teeth to resistance to air pollution). Lactose tolerance demonstrates the ability of human biology to adapt to new diets. (...) Am I really supposed to believe that there aren't genes floating around that wheat (domesticated 10ky ago) is good for?

To reiterate my point about dietary arguments, this is the sort of thing you usually cannot do with evolutionary arguments. The most what you said here should do is to slightly reduce the amount of plausible harm done by agricultural (not industrial!) foods.

It ignores wheat though, and wheat's significant disadvantages are as of now quite clearly and forcefully established (see e.g. this).

Wheat was introduced to China 3.5 kya, and it's not clear to me that it ever became a staple. I'm sure that's not the only arrow in the quiver against wheat, but whether or not it's good for the Chinese is not data I find compelling. This suggests to me that I should stick to my traditional diet as well as that wheat might be bad, rather than exclusively the second.

It's also mentioned in a comment there that fermentation might reduce the detrimental effects of wheat, so I should point out that my bread is all sourdough thanks to the dough preparation method I use.

Do you have a comprehensive blood test?

No. Where do I get one, and what am I looking for? My primary care physician?

Wheat was introduced to China 3.5 kya, and it's not clear to me that it ever became a staple.

Eh? I was taught in my classes that wheat was the staple crop of northern China in the same way that rice was the staple crop of southern China, and this was one of the distinguishing traits of the two general regions. And googling "China wheat" I find links like where there is serious international and Chinese concern over problems with the large Chinese wheat crop.

I was linking northern China with sorghum and millet, as I recalled those as original crops domesticated there. I shouldn't have been surprised that wheat eventually dominated there, though, so thanks for pointing that out.

This isn't necessarily supporting Paleo/bulletproof diet, but it is worth noting that every study showing the "health benefits" of whole wheat only compare whole wheat to refined wheat, not compared to no wheat. The results of these studies, that it reduces diabetes and various cancers, is consistent with the Paleo view as well.

So, if nothing else it would be best to cut down on non-whole wheat grains.

Might I ask how old you are? The reason I ask is that I've encountered quite a number of cases of people who abuse the living bajeezis out of their bodies but don't show any significant symptoms for many years due to sheer youth. The kind of damage that accumulates in those cases is often more subtle than blood work necessarily shows. If you're in your late 20s or so, this is less likely to be a factor.

We already know that genetic differences can affect how people respond to medicines, to the same likely applies to diets. Maybe high-wheat is good for you, but not others.

Right- that's my premise. But I'd like that to be a testable proposition, not a rationalization. An obvious step is ensuring I don't have a wheat allergy, but that seems incomplete and unsatisfying. I suspect the odds I do are so low it's barely worth checking.

I want your bread recipe, please.

First recipe in this book. It taught me pretty much everything I know about baking. (edit: in case it matters, I use King Arthur bread flour. I have not compared it to other kinds, so I can't make an informed recommendation.)

I have compared it to store-brand flour, and to Hodgson Mill. King Arthur Flour seems significantly easier to work with.

Thanks for the data!

[-][anonymous]10y 1

I'm not sure where this data will fit in with your taste preferences, but I have been taking this advice to heart lately and it has been working well for me. In particular, I have started eating more peanut butter and nuts, as well as beans and rice. At first it was a little tough, but after adjusting, I find that supplementing in between meals with lower glycemic index snacks, especially peanut butter, causes me to feel less hungry at meal times and less desiring of snacks. I've even felt less hungry from simply adding a packet of zero-calorie sweetener, such as Truvia, to a cup of tea in the afternoon (the reason for this sugar signal is explained in the link above).

I should also mention I am a vegetarian, so I don't know how meat affects this diet. I eat a lot of Indian food because it enables me to add tastes to rice that I enjoy, and thereby I consume more subjective taste experience for fewer calories. I'm still experimenting and learning, but perhaps there's something in there that can benefit you too.

I typically only snack when I've stayed up late enough that I should be eating a fourth meal, and I am infrequently hungry (the main exception to this is when I try and eat some sweet potato in between meals). When I get hungry, it's typically because I like what I'm doing too much to stop it so I can eat, and when I eat I tend to eat to fullness (and sometimes a bit beyond).

I should probably also mention that fat reduction is not a high priority for me. My BMI suggests my body fat percentage is around 15%, and the only real reason I have to push it lower is because at below 10% abs start to really pop out (and I find that rather hot).

[-][anonymous]10y 0

This is only my first-glance opinion, and it may not have much to do with goals, but my understanding is that you're a little concerned about the fact that you eat such a large amount of grains and high glycemic index foods. Eating so much bread and sweet potatoes would (again, just my guess, not a professional nutritionist) increase risk of diabetes greatly. If your goal is to reduce your bread / sweet potato consumption, and the main reason you eat those items is because of their taste and not their nutritional impact, then I think the link from Seth Roberts is very applicable. Note that the approach described there can work equally well for situations where a person wants to lose weight, gain weight, gain muscle, or train themselves to curb the cravings associated with certain foods.

The trick is to achieve whatever flavor experience you want without the same distribution of calories. My suggestion would be to eat peanut butter with your homemade bread, which should cause you to feel less hungry later and presumably help you eat less bread in general without feeling like you're not getting enough of the bread taste that you like. I don't think this has a lot to do with weight set points, unless you intend to use a diet for that type of change.

Sweet potatoes are low glycemic. This surprised me, considering how sweet they are, but they don't leave me feeling as though I've eaten a lot of sugar.

On the other hand, they might be high glycemic or something like it for a few people.

[-][anonymous]10y 0

Thanks for the links. That surprised me as well. Better update my map to match the territory!

Am I reading this right, you're eating 1-2 loaves of white bread per day?

Yes- I probably average a pound of bread a day, but I haven't done any sort of measurement of my intake. I guess that's the first place to start. So, starting tomorrow, food diary!

Even if you're not seeing a negative effect on your health I'd be worried about eating such a large quantity of something not typically regarded as healthy.

Hence this post- but I'm much less interested in general-purpose anxiety than I am in being able to check specific things. I know people think I'm in a coal mine, but what'll help me are canaries.

the second page I looked at began by talking about qi. I would be extremely skeptical of this website.

I strongly prefer wheat to rice, and rice to corn (also, white wheat to whole wheat). I enjoy the bread I bake enough to eat until I physically can't store more food in my stomach. When I bake bread with 75% wheat and 25% amaranth, it tastes noticeably worse (which I can tell because I eat what I cut and don't try to eat it all).

From personal experience, I can say that wheat is a) very addictive for me (I am much more likely to pig out on large quantities, to get cravings for the texture and flavour of wheat-based products, etc), and b) is definitely bad for me. I lack the willpower to cut it out entirely; it's quite inconvenient trying to get enough calories to sustain my very active lifestyle without wheat, especially while trying to save money by not eating much meat, and it's nearly impossible to eat out cheaply 'on the fly' wheat-free. I spend most of my life being a bit bloated and having intermittent diarrhea.

Based on anecdotal evidence from family members and friends, a lot of people feel better when they cut wheat out of their diet, even if they didn't realize it was causing problems beforehand. This may not be everyone, though...if you really feel fine, then you're most likely fine with it.

The main thing that worries me about your stated diet is protein intake. Unless you supplement with a lot of eggs and dairy, grains contain incomplete protein, i.e. not enough of certain essential amino acids, which means that your body can't use the amino acids to build tissue proteins at all and just breaks them down for energy. Do you eat any beans, nuts, etc? Legumes combined with grain create complete protein, not quite as high-quality as meat sources, but almost. And beans are cheap, and come with the added benefit of high fibre.

If you want a small change in a paleo direction, put butter or olive oil on your bread.

Sometimes I put olive oil in my bread, since it works better as pizza dough. Never was a fan of the whole "dip your bread in oil" thing, but I suppose I could give it a few more tries.

The impression I get, though, is that if I'm going to see major benefits from switching to paleo those benefits will come more from excluding grains (and going into ketosis) than from including more fats.

Dip your bread in oil with red pepper flakes, crushed rosemary, roasted garlic, and a little salt.

Alternative: balsamic vinegar or lemon juice.

Is that to improve the flavor (which I think it would do very nicely) or for health benefits?

Flavor. I have no relevant expertise on the health benefits, but I think all the listed items except the salt are neutral or better.

High quality olive oil tastes much better than the cheapest oil (even within the EVOO category), which matters a lot for dipping. Flavors differ a lot by variety: there are buttery, olivey, "fresh" herbal-tasting, and peppery oils (that's within the spectrum of unflavored oils, obviously flavored ones differ along more dimensions).

The impression I get, though, is that if I'm going to see major benefits from switching to paleo those benefits will come more from excluding grains (and going into ketosis) than from including more fats

I'd say the opposite. There's quite a bit of evidence that eating certain types of fats is good for your brain. I personally have found significant mental performance after eating butter or MCT oil.

Ketosis is a bit extreme in excluding carbs (especially for someone who lives on bread!) and isn't necessary. I couldn't measure any difference in mental performance in/out of ketosis (though there could have been subtle differences in a dimension I wasn't measuring. I generally stay out of ketosis because when I'm in ketosis I won't eat enough which can effect my energy levels until I force myself to eat.

Hm. Helpful! I'm guessing I should set up a simple arithmetic test like he has so I can track mental changes that way, then try adding some butter to my diet.

I've been using Alexei's life tracking app, which is quite convenient for this sort of thing, if you have a droid.

Be careful with what you're testing, and be mindful when you're performing cognitively demanding tests. Simple arithmetic (e.g. 5+6) for the most part is a single memory look up, not actually doing math. Times for me did not really suffer much under severe sleep deprivation, though my overall performance did. MCT oil did show improvements here (by the time I started jotting numbers down I had fully incorperated dairy fat so it's harder to test). You might want to try dual N back or something to test other dimensions.

Playing guitar has been a pretty sensitive cognitive ability sensor. Piracetam makes me noticeably faster (even to friends who see me play less than monthly), but butter doesn't seem to speed up my absolute speed much. What it does is help me reach my top speed with sequences of notes that I'm not quite as used to playing. I'd guess that single look up arithmetic would under report the awesomeness of butter.

Agreed that it's a simple memory look up. I think the benefit of it over dual n-back (something I've started but have only done a little bit) is that the practice effects are much smaller, since I've already heavily practiced doing simple arithmetic, whereas results with dual n-back will also show my normal learning.

But, I suppose if I'm also interested in learning speed (and I am) and not just memory speed, then that makes DNB a better choice.

I've got experience with serious sleep dep and have a PVT program already ready to go, but that's really boring to use and I don't expect it would give any interesting results when I add fat to my diet.

There are definitely benefits of a simple arithmetic test, I just didn't want you to underestimate the benefits by measuring the wrong thing.

Tell me more about your PVT program. I should probably try something like that.

It's nothing special- it runs in Ruby Shoes and I can email you the source code if it would be at all interesting. I wrote it after reading a number of papers describing sleep dep experiments that used it.

It shows a stimulus (increasing time in red in the center of a black background), then once you click in the upper left corner it waits 2-10 seconds (chosen at random) and then shows you another stimulus. This goes on for ten minutes. You get a host of reaction times (and so can watch that change) but the real value (and the reason it's boring) is that your ability to sustain concentration drops with sleep deprivation, and so lapses (where it takes you at least twice as long as your baseline, typically .5 to .8 seconds) start showing up. They become more frequent and longer the more sleep deprived you get (at my worst, it took me about fifteen seconds to respond to a stimulus; I didn't code it but the paper I took the parameters from had a buzzer sound if you went 30 seconds without a response). It outputs the times (absolute times of stimuli, responses, and false starts) to a text file, and I never wrote any analysis code because I considered the affiliated sleep experiment a failure.

It's also a very good proxy for driving- it involves a lot of staring at boring, barely changing views, and then when one gets into an accident, it's typically something that noticing it coming a half-second earlier could have helped a lot with. And so when you find that a half-second of inattention goes from a <1% of the time thing to a 5% of the time thing, you swear off the heavy machinery.

easy change: less wheat, more beans and rice.

Right, but that's a significant (for me) decline in taste. I'd like to be sure I'm getting something back; what should I expect to see if that's helping?

if you want short term gains I'm not sure what nutrition advice I can give.

For me eating healthy is about living longer in a more optimal state of health, short term I can pump in garbage and not see a huge change because I'm young.