Kurt Harris MD

An Archevore is someone who eats based on essential principles, and also someone who hungers for essential principles. Take your pick.

Exploring these principles is one of my interests, but not the only one.

So you may find commentary here about other issues in medicine, health, other sciences, or just about anything.

Feed The Archevore

Archevore is written, produced, and directed by me. I am an independent science writer with no outside sponsorship from any private firm, NGO or, Zeus forbid, government agency. Donations are greatly appreciated. 


In addition to buying from the book list, you can also support Archevore by making all of your Amazon purchases for any item through the Amazon Portal below.

Amazon Portal

 

 

Buy gold online - quickly, safely and at low prices

 

Wednesday
Feb022011

How to Read This Blog

I like to think that this blog functions a bit like a photo album or personal journal. Each posting is a snapshot into what I was thinking at the time it was written.

I don't post a lot of music videos or any of my poetry. I also don't post much about many other deep interests I have like economics, investing, the brain and consciousness, general philosophy of science, etc.

But it's still pretty eclectic in types of posts.

Addressing a couple of comments and questions I've had recently:

Question -You mentioned that you have changed your views on x recently - are you going to go back and revise all of your of previous blog posts and 1000 or so comments to reflect that?

Answer - No. To use the photo album metaphor, this is like saying "I see you used to have more hair 3 years ago. I think you should go back and photoshop the older images so they look more like you now." Or, "Your poetry is so much better now than when you started. Why don't you re-write all your older pieces"? All the blogs I admire seem similar in this regard. You can read old posts of Peter's where he is taking 10,000 units a day of D3, but now he takes zero. Stephan initially entertained the idea that low total carbohydrate might be necessary for health. He no longer believes that, and although he has commented on that change of opinion, he has not gone back and deleted or modified old blogs posts, as far as I can tell.

So please do not use this blog as if it were a dietary on-line Harrison's Internal Medicine - continually updated for accuracy. Like every other blog of this type - a journal essentially - the most recent posts will be closest to my current thinking. The exception to this is "get started" which I strive to keep updated from time to time. So you can use the blog however you like, but if you are new you might first read "get started" and then read in reverse chronological order. That is what I do when I find a new blog I like. Going in reverse is good as you can immediately identify what is out of date as you encounter it in the older posts. I am trying to do this with John Hawks weblog right now, for instance. Learning a lot there!

Question: How come you don't provide more references?

Answer: The photo album is full of different photographs. Spontaneous action photos, artsy black and whites, panoramic wide angle landscapes, crappy snapshots...

There are some posts that are explications of peer reviewed articles that will obviously reference the article referred to. There are some in-depth posts that have a few supporting references where the reference is critical to what I am saying, or where the point is especially controversial. Some are short comments in response to a question. Maybe I have a preliminary opinion on something, but have not assembled all the references in one easy-to-access spot.

The main species of post is the essay. These are editorials. They are opinion pieces about dietary science that are informed by the reading and digestion of literally thousands of abstracts, perhaps 100 books and close to a thousand full text papers per year. When I write a series like "No such thing as a macronutrient" it represents about 10 hours of actual typing preceded by days of re-reading various references, and synthesis of ideas and knowledge gleaned over years. To add all the "references" that support every claim or opinion expressed would take days - and for the "macronutrient" essay there would easily be over 100 references.

To use the "macronutrient" example, you can do a lot of the fact checking yourself. It's really easy. Just go to pubmed or google scholar or google and type in "fructose" and "insulin resistance" or "fatty liver". 

Some day I will start the first of several books. I can write pretty fast, so most of the work will involve compiling references, fact-checking, getting professional editing, etc. I could do that level of detail for each essay-length blog post, I suppose, but then I am basically providing a free on-line book in installments, the blog will get even slower, and the book would either be redundant or never get written.

Think of it like this. Let's say you invited someone to coffee to talk about their field of expertise. Let's use John Hawks as example. That's a good example for me because I could learn a lot from John Hawks. He has spent his career researching and teaching paleoanthropology. If I asked him to coffee so I could pick his brain for an hour, I don't think I would be disappointed if he failed to bring his laptop loaded with PDFs of all the papers he's ever read. After asking his take on "Out of Africa" I would not stop him mid-sentence and demand to know which scholarly articles he can cite in support of his ideas.  I would just let him tell the story as a narrative and see what I could learn, and I would trust him to mention critical references if and when he chose to do so.

Now, I am not suggesting or demanding that you grant me that level of trust or authority. I am saying, though, that those are the rules. I am just publishing a journal of things I think about - opinions about scientific ideas and hypotheses, and you are free to heed or discount whatever I say as you see fit.

But it's really just a chat at the coffehouse or a browse through someone's photo album or a poetry journal.

It's not the New England Journal of Medicine.

Finally, I love writing blog posts and I enjoy reading papers and books. I despise spending extra precious time on the computer compiling references and creating links. That is just work, frankly, and if I was forced to a "review article" standard I would stop blogging altogether and just play my guitar and do more drawing and painting.

Question: You said you welcomed friendly emails. I have a special dietary issue. Can you give me advice?

I am very happy if you find my writings useful or entertaining. I have absolutely no time to do one-on-one dietary coaching. So I must clarify, "friendly" does not mean any email that contains a request that would need a response. 

Saturday
Jan292011

N-3 supplementation recommendations

I started the current series conflicted about how much detail to put into what I see as mostly an essay on nutritional methodism or philosophy. I could have made it shorter with links, but wanted it to be able to stand alone. So I decided to write it long but break it up into parts. So there is some substance there as well. I am trying to lay the groundwork for future posts that are more in-depth and dealing with the NAD -the putative neolithic agents of disease. 

Chris Kresser of the excellent blog Healthy Skeptic has read Part I and decided to throw down the gauntlet on the issue of n-3 supplementation. You can see his full remarks in the comments section of the previous post. I've abstracted his questions in italic and answered in Roman.

Forgot to mention this, but Beth reminded me. In your article you mentioned supplementing with omega-3 when omega-6 is unavoidably high. What constitutes "unavoidable" for you? Because I think you'd agree that dramatically reducing n-6 intake is a much better way to balance the n-3:n-6 ratio than maintaining n-6 intake at current levels and taking 20g/d of fish oil.

......I'd be curious to know where you land this, Kurt. It's a pretty controversial issue in the Paleo world.

The six people who have read everything on the blog and all my comments elsewhere know that I favor reduction of n-6 in the diet as the way to correct the 6:3 ratio and that I am skeptical of supplementation with fish oil. I would never advocate fish oil at the level of 20g/day and even though I have great affection and respect for Robb Wolf and Dr. BG, I do not think using a "fish oil calculator" that recommends tablespoon or multigram quantities is a great idea. 

At the same time, I find it hard to imagine how 1-2 g/day of known-to-be fresh fish oil or n-3s in tinned sardines is metabolically much different than getting a bit more n-3 by buying pastured beef or butter. As I eat both pastured beef and butter and occasional fish, I only take the odd teaspoon or so of Carlson's fish oil now and then, and I probably would not need to do that all.

The "unavoidably high" group I am referring to is your relative who will make zero change in their overall diet but might be willing to take a few fish oil pills along with their BP meds and useless multivitamins and neutraceuticals, etc.

For the serious readers of my blog, I recommend a high GRAF, low TemPO, and even low NRAF diet at steps one two and three to optimize the 6:3 ratio and minimize total PUFA. The point is BOTH the ratio an the total are important, so fix them BOTH.

I can see the logic of higher doses of dietary n-3 on the theory that one might accelerate the displacement of the unwanted n-6 in cell membranes. But although I don't know the pharmacokinetics of this maneuver inside and out, I suspect the cells "see" serum levels of n-3 and n-6, but not necessarily in the ratio we eat it! So my cartoon concept of this is that 15 g /day DHA/EPA may bump the serum or lipoprotein borne level only slightly (homeostasis and all suggests the serum level might be regulated - we do it for other fatty acids), at the same time there is an order of magnitude more overall PUFA (with n-3 more unstable than n-6) at the level of the gut and liver. It seems we might preferentially burn the excess, but perhaps we store it all and it speeds the displacement of N-6 from the cell membranes. I kind of doubt that, though.

I prefer to avoid all this, and to allow n-6 to just wash out, even if it takes 2 years.

This bias is not something I can prove just with literature quotation, as there are no trials I am aware of that compare n-6 reduction over time to correcting the 6:3 ratio at the expense of increasing total PUFA to a level evan further outside the EM2 than it is already. I think the burden of proof is on those recommending supplementation for this reason. The burden of proof must be high when recommending anything we definitely did not evolve with. I have not read everything on this - who has?, but I've read enough to eat this way in my own family.

 So start your high GRAF, low TemPO, low NRAF diet now and hold off on the fish pills. You won't need to calculate anything then, and with time your cell membranes will have a 6:3 ratio that mimics the EM2.

 

For more info, you can read Fish Oil or not? and Fats and Oils.

Saturday
Jan292011

There is No Such Thing as a Macronutrient Part I - Fats

What is a macronutrient?

Macro means large, in this case large in scale or quantity. In common usage, a macronutrient is a class of nutrient that is required in larger amounts – larger amounts than micronutrients like vitamins and trace minerals.

Sometimes the larger scale minerals - sodium, magnesium and calcium - are considered macronutrients or macrominerals. But usually we consider the caloric and structural molecules we need in large amount to be “the macronutrients”. They have for many years been classified by their chemical structure into three groups.

Fats        

triacylglycerols  - three fatty acids on a glycerol backbone. Used as fuel and for structure and function. Long chain saturated fatty acids (LCSFA) in particular are very energy dense, like diesel fuel for the body.

Carbohydrates        

composed of simple sugars (CH2O)n, n>=3. Used as fuel and for structure.

Proteins

Polypeptides – polymers composed of combinations of 20 different amino acids with highly variable side chains – Used for structure, function (enzymes) and for fuel as a back-up. The variability of the amino acid side chains makes for an almost infinite variety of protein shapes and functions – from a spider’s web stronger than steel to enzymes with highly specific functions.

For more than a hundred years, diets have been discussed with regard to their ratio of macronutrients, that is, whether they are composed of relatively more or fewer Carbs, Fat or Protein. In 2011, we are still arguing about the merit of diets that are called “high carb” or “low fat” or “low protein”, as if this was a very important, if not the most important dietary parameter.

Indeed C% + F% + Pro% = 100%

By definition.

But so what?

This is appealing at a crude level, as certainly for parts of modern history, and in researching metabolism, it is interesting to learn how we can manipulate macronutrient ratios and what effect this has, especially at the extremes. The effects we see from manipulating them, like with a low carbohydrate diet, certainly reinforce the impression that macronutrient ratios are very important. And the lipid hypothesis, both to those who buy it and those who think it’s nonsense, forces us to take positions on either side of a “low fat/high fat” axis.

But I would like to suggest that the whole concept of macronutrient, like that of the calorie, is determining our language game in such a way that the conversation is not making much sense.

I think we are stuck in a 20th century dietary paradigmatic purgatory, and we could move to a more fruitful one by speaking less of macronutrients, or at least coming up with a new taxonomy for them.

My reading of paleoanthropology and ethology of modern hunter-gatherers suggests that it may have been possible to eat healthy and avoid diseases of civilization (including accidental weight gain) with no concept whatever of macronutrients. Are there words for protein or fat or carbohydrate amongst the Hadza or the !Kung San or the Native Sioux or the Kitavans in their native languages? I am sure there are words for fish and cassava and meat and such. Do they have discussions about how to apportion their calories among these? And how did they and every other animal on the planet and the entire ancestral hominid lineage manage without calculators and scales?

Approaching from the other end, looking at what we know about food science with just moderate sophistication, of what use is the concept of "macronutrient" now?

Fats

Let’s look at fats first.

Is “high fat” good or bad? Are “fats” good, bad or neutral?

Don’t we know enough now to ask “which fats”?

Under the rubric of “fat” we have the perfect human fuels – what I call #1 Diesel – long chain saturated fatty acids (LCSFA) like Myristic (C14), palmitic (C16) and Stearic (C18) acid as found in the milk and meat of ruminants. These are, by no coincidence, the same saturated fatty acids we humans use to store energy in our own bodies – our own internal #1 Diesel. As far as I can tell, there is no limit to the amount of this kind of fat that you can eat, as long as your protein and micronutrient needs are being met. You cannot eat “too much” saturated fat unless you are not getting enough of something else. This is just the principle of displacement, though, and has nothing to do with any toxicity of these fats as food. Of course, these LCSFAs are the very fats that, in the bizarro-world of conventional nutritional wisdom, are supposedly the most dangerous and cause heart disease and cancer. That there is no real evidence for this belief, and that LCSFA are actually one of the most healthful sources of caloric fuel, is what makes embracing them and rejecting the lipid hypothesis the lodestar of PaNu or any rational approach to diet. You can’t take the most important step in improving your diet until you do this.  

Stearic Acid – #1 Diesel  for the human body

 

Sidebar:

An incommensurable dilemma presents a binary choice where both choices cannot be correct. It is often the crux of establishing a new paradigm. A good example of an incommensurable is the Ptolemaic vs the heliocentric model of the solar system. Not all incommensurable dilemmas are solvable by accepting one or the other paradigm. Sometimes the question is framed wrong or the language is defective – that is what I am suggesting in the case of “low fat vs low carb” – the framework of the dilemma itself is wrong – so to move to a more workable paradigm requires asking a different question with new language.

In the larger case of how to eat healthy in a manner supported by medicine and evolutionary science, the truth value of the lipid hypothesis is the incommensurable dilemma. You simply have to choose whether it is true or not. You really cannot avoid the question. If you hedge, you end up running into nonsense like being “low fat” and “low carb” at the same time. You may end up terrified of both saturated fat and starch and then you’ve made no progress at all, as you really should not be afraid of either one*!

I believe dealing with the incommensurable dilemma of the lipid hypothesis is far more important than using any kind of evolutionary reasoning or being at all “paleo”. If you can’t make the crux move, with all there is riding on it culturally (all the terrible public health consequences), and all the evidence laid out for you, there is no point in going in more speculative directions, like “what did paleo man eat?”. This is partly why “paleo” on my blog now means just old or ancient, instead of Paleolithic – with all the speculative and unknowable baggage that reference to the Paleolithic carries.

Next we have so-called medium chain triglycerides (MCT). These are saturated fats with fatty acids that have a chain length of 6 to 12. They are found in coconut and mother’s milk and have some interesting properties. Even though they are saturated and quite similar to LCSFAs, they are metabolized differently. They do not require bile acids for digestion, and they go directly to the liver via the portal vein. They seem to undergo obligate metabolic processing that results in ketone bodies – so they will enhance the level of ketosis (ketones in the blood) even without eating a VLC ketogenic diet. They can be good fuel sources, but in addition may prove useful in a therapeutic context by enhancing ketosis.

Next, consider n-6 polyunsaturated fatty acids or n-6 PUFA. The principal of these is Linoleic acid – the Third Horseman (of the dietary apocalypse, one of our three Neolithic Agents of Disease).

The formula is 18:2 n-6. There are two unsaturated, reactive, oxidizable double bonds, with the first found at the 6 position. Linoliec acid (LA) is an essential fatty acid. We must have some in our diets, as we can’t make it ourselves. However, since the introduction of plant and seed oils into our diets as food as a by-product of industrial extraction techniques that were impossible before the machine age, the western diet has seen quantities of LA that are often nearly an order of magnitude greater than what humans or hominin ancestors could have ever experienced.  A healthy level of LA might be as much as 2-3% of total calories. Intakes in north america on the SAD may be as high as 10% or more. Excess linoleic acid alters the tissue 6:3 ratio which in turn affects inflammation and the immune system, encourages a thrombotic state, may cause cancer, may interfere with leptin signaling in the brain, may affect gut permeability, may be hepatotoxic, may contribute to oxidative damage affecting the development of atherosclerosis and through a variety of mechanisms, including affecting insulin sensitivity, may contribute to metabolic syndrome and obesity. The amounts in our diet are far outside of any reasonably likely evolutionary experience and even though LA is essential and LCSFAs are not, it would be fair to consider LA a metabolic poison when getting 10% of calories from it. A bad fat indeed.

Again, we should note that many standard bearers of the reigning paradigm consider LA a “good fat” - having almost medicinal properties against heart disease. As there is no direct evidence for this, their belief only makes sense once you realize that this is because replacing LCSFAs in the diet with LA “improves” LDL numbers. So their faith in the lipid hypothesis essentially compels them to believe that the worst naturally occurring fat you can eat is a “good fat”. Do you see now why confronting the lipid hypothesis is your crux move in sorting out the science of diet?

The next fats to consider are the long chain n-3 PUFAs Docosahexanoic acid (DHA) and Eicosapentanoic acid (EPA). These are polyunsaturated fats that are only obtainable in quantity in animal products like fish and grass-finished ruminants.

Sidebar:

The 6:3 ratio in plants should be ignored. Nuts or seeds that have any n-3 at all should be thought of only as sources of unwanted PUFA. The n-3 they contain is not converted to the useful long chain n-3s we need at a high enough rate to even consider them food. Plant n-3s like alpha-linolenic acid (ALA) from flaxseed oil should just be considered sources of excess total PUFA like LA, only even more likely to cause oxidative damage as they have more double bonds and are less heat-stable! Plant oils like linseed oil (contains ALA) are literally varnish. I use them to make furniture. I use them for oil paintings as a medium. I avoid eating them.

DHA/EPA are found in pastured butter, beef and lamb, and fatty fish. These are the n-3s that balance the n-6s and that determine the denominator of our 6:3 ratio. As I’ve said before, your total PUFA should be no more than 4% of calories and your 6:3 ratio no higher than 3:1, preferably less than 2:1. You will not achieve this if you eat a lot of chicken or handfuls of nuts. If a person has unavoidably high n-6 in the diet, then adding some DHA/EPA is a good idea. The higher the total PUFA above 4%, the more benefit to supplemental EPA/DHA. So n-3 PUFA is a good fat if you need it, but once your 6:3 ratio is OK, it becomes a bad fat, as it is even more unstable and reactive in your body than n-6 LA, as it has more double bonds. So here we have a fat whose goodness is entirely contextual, depending on the rest of the diet (consumption of n-6) and even the present status of tissue saturation by n-6 due to dietary history.

Note: One might accelerate the normalization of the 6:3 ratio by supplementing beyond the calculated requirement for a year or so, but I would not exceed the required ratio by more than an extra gram or so of DHA/EPA combined.

Finally, we can consider the dreaded trans fats. A trans fat is a triglyceride where the fatty acid is unsaturated – it has at least one double bond, and the carbon atoms on either side of the double bond have hydrogen atoms on opposite sides. This makes the shape of the molecule similar to a saturated fat, so that it is more-or-less straight, like this:

  

Elaidic Acid

This is Elaidic acid, the principle trans fat in hydrogenated veggie oils.

 

Oleic Acid

For comparison, here is oleic acid, which is found in beef fat and olive oil, and is considered to be relatively healthy (by those on both sides of the lipid hypothesis divide, interestingly enough). The only difference is the orientation of the hydrogen atoms that were added, and this results in a bend for oleic acid versus the straighter configuration of Elaidic acid.

The effect the bend has is to lower the melting point of the fat, as the bend makes it harder for the molecules to fit tightly together where Van Der Waals forces between the atoms make the structure more crystalline or solid.

Now we should make it clear that there are trans fats found in nature. They can be up to a few percent (usually 2%) of the fat in the meat of ruminants and in dairy fats. However, the predominant type of natural trans fats are conjugated linoleic acid (CLA), otherwise known as rumenic acid, and its precursor vaccenic acid. Vaccenic acid (C18:1 trans-11) can be converted to CLA by humans, which is not only proposed likely free of the negative metabolic effects of the dominant artificial trans fats (like Elaidic acid), but possibly has anti-cancer properties. Both of these “good trans fats” are made by bacteria in the stomach of ruminants, and like the long chain n-3 PUFAs we like, they are made in greater amounts on a grass-only diet. So grass-finished ruminant milk products and fatty meat will have these good trans fats, and they are good for us.

So why are there artificial trans fats and why are they bad?

This Wikipedia article gives a good account of the history, but briefly, artificial trans fats were first introduced as a result of attempts to process waste industrial vegetable oils (soy, cotton seed) into something resembling animal fats like butter and lard. Because oils rich in n-6 are rather soupy at room temperature (those kinks from the -cis configuration), they can be hydrogenated in order to behave more like saturated animal fats in the kitchen and in baking. This is achieved by adding hydrogen at the double bonds – hydrogenation. If there is complete hydrogenation – say oleic acid is completely saturated to C18 stearic acid – this is not a problem as you have just made a tasty saturated fat. But partial hydrogenation – say, turning a PUFA into a MUFA with a single double bond – will create these artificial trans fats. The catalyst for the reaction is indiscriminate and will create roughly half –cis configuration and roughly half –trans. The cis are not likely a problem, but -trans fats resulting from hydrogenation of something that started out chock-full of linoleic acid -like cottonseed or soy or corn oil - will go from 0 to 25% or more.

Artificial trans fats have been linked to variety of metabolic disturbances. Read Mary Enig’s book Know Your Fats for a good summary. Some of these are based on effects on supposed surrogate markers for atherosclerosis like LDL and HDL, which of course makes me a bit skeptical, but even with no direct human evidence of mortality, they should not be eaten, as they have simply never been part of the human diet. (I am perfectly comfortable proscribing substances that are literally not food, even on weak evidence)

Before the beginning of the 20th century, butter and lard were king. After mid century, replacement of animal fats with cheap industrial oils and margarines meant our diets had almost as much artificial trans fats as saturated fats in them. Remember this when looking at literature indicting saturated fat, as for much of the 20th century they were often lumped together in studies looking at fat in the diet.

Nowadays, trans fats are of less concern, as they are limited by law to no more than 7% of total fats by removing the -trans isomers and leaving the -cis ones, and even Crisco is now relatively low in them. The easy way to avoid them is to totally avoid anything with “partially hydrogenated vegetable oil” in it – even if they claim a low trans fat content. Better yet, completely avoid all industrial vegetable oils –they all have too much n-6, even olive oil.

So in summary, we have added two more subclasses of fats, natural trans fats –naturally occurring vaccenic and rumenic acid, and artificial trans fats – chiefly Elaidic acid. One class is very good and the other is very bad.

We can tally these up now and even create some more food-relevant subcategories:

A New fat Taxonomy

LCSFA - long chain saturated fatty acids - best

Monunsaturates - Oleic acid from animal fats and olive oil - good

n-6 PUFA - Technically essential but in huge excess due to technology - bad

n-3 PUFA - necessary to balance excess n-6 but otherwise bad - contextual

MCT - Medium chain saturated fats - good in reasonable amounts

NTF - Natural trans fats like vaccenic and rumenic acid - good

ATF - Artificial trans fats like Elaidic acid -not found in nature - bad

The preceding would be the chemically based, "nutritionist" taxonomy. But we can construct a real-food based taxonomy by recombining the nutritionist elements into a 6 part scheme like this...

GRAF

Grass finished ruminant animal fats. LCSFA and Mono in roughly equal amounts. Same small amounts of n-6 as IRAF, but more n-3 to balance. More NTF. Best.

IRAF

 Industrial grain-fed ruminant animal fats - LCSFA and Monounsaturates in roughly equal amounts with smaller amounts of PUFA - may be deficient in n-3 PUFA. Has less NTF. good

NRAF

Non-ruminant animal fat. Highly variable. Has LCSFA but may be very high in n-6 and if grain fed poor in n-3. More sensitive to diet of the animal than IRAF or GRAF. fair if pastured but poor otherwise. Think factory chicken.

TemPO

Temperate plant oils. Artificially abundant due to technology. excess n-6, effectively no n- 3. Some MUFA. Little LCSFA bad

TroPO

Tropical Plant Oils. Coconut and Palm oils. Good source of MCT and/or LCSFA - lower in n-6 and n-3. These are also made available by processing technology, but their content is better than TemPO. Good.

FF

Frankenfats. TemPO chemically modified by hydrogenation. Very bad - avoid completely.

 

I hope I have convinced you that the macronutrient label of "fat" as in "high fat" is or "low fat" diet is metabolically not very useful. Keep this in mind when you read anything at all in the literature about "fat".

Perhaps you can help me change the nutritional language game by using my new fat taxonomy and resisting the old one, even if you disagree with my judgements about their nutritional merit.

When you discuss diets or ways of eating, you can use language like "I eat PaNu, a high GRAF, low TemPO diet".

Or, "PaNu is a high GRAF, low TemPO, low NRAF, zero FF way of eating". PaNu is "low fat" in regard to TemPO and n-6 PUFA, but very high fat in regard to LCSFA, Mono and GRAF.

Or "Dr. Davis advocates a diet free of wheat, but unlike PaNu it is low GRAF and IRAF, but high TemPO and high n-6 PUFA". Davis is "low fat" in terms of animal fats, but "very high fat" in terms of n-6 PUFA.

We'll deal with "Carbohydrates" and offer a new taxonomy for them in Part II.


Thursday
Jan272011

More N = 1

Reader Dallas Hartwig makes some comments in italics, and my responses are in roman.

Thanks for yet another thought-provoking article. I appreciated the cordial-ish exchange on Robb's site, and (unlike some other commenters there) I'm completely good with simple disagreement. Shoot, why does everyone need to be "right"? I have a modest body of experience and education, and both of those are constantly growing. And, like other folks that you've posted about, my position on some things sometimes shifts over time as my understanding and experience changes me. On other things, my current position stays firmly where it has been for years. Anyway, on to my actual point.

I think that there should be a marriage of peer-reviewed science and personal/collective experience.

You say that as if it is something I might disagree with. I would not disagree.

That partnership allows for real-time improvements, instead of waiting for the consensus of published, peer-reviewed research to indict or acquit one particular food or food group.

I think where we disagree is on how robust the evidence should be before an athletic trainer or a physician tells someone that as a rule they can no longer eat a whole category of food. I feel the case against wheat is an order of magnitude more convincing than for cream and butter and non-soy legumes. So I would be loathe to start with that a proscription, just like I am loathe to say don't eat beef because some people have a beef allergy (as discussed on the thread at Robb's) I am not saying don't pre-judge without a randomized trial, but I am saying that the case against these latter foods is not as defensible as wheat, etc.

The solo act of a science-only approach, while it avoids the (valid) placebo concern, spawns extended periods of time (decades!) with no definitive answer for the masses that are so, so sick these days. I agree with you on the general reduction (or complete omission) of wheat/gluten, fructose, and LA, but I'm not clear on how both you and I concluded that these things are less-than-awesome, since scientific research is far from "complete and conclusive" on even these topics.

I never said "complete and conclusive" nor have i advocated a solo act of only peer reviewed science- those are your words. I said the scientific reasoning should be sound and there should be evidence. And then N=1 should be applied with caution, with emphasis on objective measures of health. This is based on my 20+ years of experience as a physician, and how relying on subjective impressions can be very hazardous.

So while there are some publications that suggest that most folks would do better without these, we've "married" that incomplete scientific evidence with experience, and forged our opinion, right? I think the general dismissal of individual experience as a valuable indicator leaves the general public open to more "science-based" recommendations like the Food Pyramid, etc.

You are misunderstanding what I am saying. If you re-read it, you will see it is not a general dismissal at all. It is a cautionary warning. When we come up with something like my 12 steps or your diet rules, we are indeed making a heuristic that is similar to the food pyramid. We indeed do it on incomplete information, and so should be pretty conservative in what we proscribe - so to avoid it being rightfully dismissed as scientifically indefensible, like the food pyramid.

......I guess I feel like the complete reliance on published literature to support one's position is unhelpful for folks desperately trying to sort through all the conflicting recommendations.

Sometimes we must sacrifice helpfulness and convenience for accuracy.

Robb's "try it and see" approach was, I suspect, a response to the constant rebuttal of some scientific suggestions (if not conclusions) that he drew from his education and research.

I don't think it's at all detrimental to perform a case study of yourself by removing foods, reintroducing them, and evaluating how you do.

Thinking that this is more meaningful than it really is what is detrimental. Not the doing it, just thinking you have really proved something. Like thinking wheat is harmless if you can eat an entire pizza in one sitting and have no ill effects. I can do this no problem. OTOH, if I eat too much onion I feel ill from the FOS fermenting in my gut. Does that mean pizza is good and onions are bad*? Surely you can see the deeply flawed logic in this kind of self-experimentation.

However, I do think that waiting til all possible doubt has been removed by scientific study could have some tangible consequences. I, personally, would rather err on the prudent side, even if that's construed as a precursor to paranoia.

You already know I have never advocated removal of all possible doubt. I am a neuroradiologist. I deal with Baye's theorem and conditional probabilities for a living. And I think that messing with people diets (and consequently their heads) is serious stuff, so I prefer to err further on the side of being more sure that there is a sound case against a proscribed food.

It's not paranoia I'm worried about. It is accuracy and the long-term defensibility of what is being recommended.

Thanks for the dialogue.

You are welcome!

 

* For newbies, the answer is the pizza (wheat flour) is bad even if I feel fine, and the onion is harmless even if I am doubled over with temporary gas pain.

Thursday
Jan272011

N = 1?

In common usage in the medical world, the phrase "n=1" is often used in the pejorative. A colleague tells you of some clinical scenario, perhaps involving some patient and their favorable response to medical intervention. You doubt the universality of the observation, so you might say "Yeah, but that's just n=1". You are trying to say that it is anecdotal. That we should be more scientific in our thinking. That n=1 is obviously not enough. It is the opposite of a large, randomized well-controlled trial.

In the "paleo" and health blogosphere, I see "N=1" experimentation, used unironically as a guiding principle, almost a fetish.

It is often explicitly stated that "how it works for you" and especially "how you feel" is the ultimate arbiter of the advisability of dietary maneuvers.

Ultimately, in the literal sense, as the LAST thing to be considered, this is true.

But the starting point should always be the soundest science - based evidence and reasoning.*

Many practicing physicians will instantly understand my skepticism, as will students of human consciousness and anyone who appreciates the radical subjectivity of human experience.

I had a patient once who claimed she was allergic to every food except skinless chicken breast and oatmeal. She ate nothing else for years.

I have had patients with back pain claim that their backs were ruined by the most trivial of events, such as lifting a 10 lb UPS package 5 years previously.

One lady claimed a large swelling on her leg occurred at a result of tripping over her cat. It turned out to be a malignant tumor.

I had a male patient present with a golfball-sized tumor his testicle, and when I told he and his young wife that it was a tumor, they both swore it had only appeared a few days ago. She was actually incensed at me, insisting that she was frequently and intimately familiar with her husbands anatomy, as there is NO WAY she would not have noticed such a thing…

One of the things I am concerned about in encouraging people to closely monitor their responses to food is creating an army of people with real but essentially psychosomatic (created by the brain) reactions to possibly harmless foods, at the same time as we are reassurring them that if they “feel fine” eating something, then that means it is harmless.

I felt absolutely fine drinking up 4 coke classics a day in my youth, no lie. I was in good shape and not obese.

Conversely, a person with undiagnosed celiac disease may have no symptoms whatsoever until they have type I diabetes, Hashomoto’s thyroiditis or neuropathy or cancer.

Might the fact that you feel so much better without something in your diet be because because of your expectations?

Both the placebo effect and its opposite, the nocebo effect are very powerful. Often more powerful than actual drugs in clinical trials where the effects can be assessed.

Why should we doubt that we can fool ourselves about just about anything?

Don't some people think they feel great on whole grains and soy and maybe even vegan fare, attributing even negative symptoms to "detox" or "candida die off" or other nonsense?

Just to be nice and sparkling clear, I am not saying one should not pay attention to how you feel with dietary changes. Far from it.

I am really just taking issue with two things:

1) Using the subjective experience as “proof” that a salutary change was made. This is especially difficult when you are changing a huge number of dietary variables at the same time and, often altering the physical culture the person experiences in big ways. Arguing that if x feels better without y is highly unreliable. In the case of both wheat and “dairy”, we should make our arguments based on plausible science first, as the “feeling” may be absent or misleading and hard to separate from the other variables in either case.

2) The contradiction between telling people to “suck it up” if they are suffering without their wheat or sugar or starches or milk, at that same time that they are being told to closely monitor how they feel as a gauge to what to eat.

It will be best to always emphasize N=1 objective measures of health like blood pressure, waist size, glucoregulation, etc.

Ultimate means last. Use N=1 subjective experience as the ultimate - last - test in this literal sense.

If you apply N=1 subjective evaluation to dietary maneuvers that were not totally grounded on science in the first place, you are liable to seriously fool yourself.

*This is really the only thing differentiating sound from unsound. The intitial science behind the dietary principle is the only differentiator. Just read 30 Bananas a Day or any vegan website to see what I mean. The subjective experience in the short term cannot be the crux. Humans can eat anything in the short term.