Blog

5

Hi All,

In medical statistics, Berkson's paradox refers to the error in estimating how strong a risk factor something is for a disease because one improperly generalizes from a sick population. Dr. Harris's medical knowledge is impressive and valuable. But, he sees a biased sample of humanity. Cures for sick people aren't necessarily the basis of preventative treatment. So, what he sees work for sick people need not translate in advice for those who want to maintain rather than recover health.

To push the envelope, why is his experience helping to diagnose the sick relevant to preventative medicine like proper nutrition?

In sum, I'd be interested in the community's thoughts on how Paleo approaches to treating disease differ from maintaining health. In particular, other than for parsimony's sake, why should we assume them to be the same, when evidence exists that they aren't? For example, the steps to prevent coronary artery occlusion will lead to death after a certain degree of occlusion. (At the risk of a reduction to absurdity, exercising to prevent a bad ticker is a good idea, exercising with a sufficiently bad ticker is a bad idea.)

Thanks in advance for your thoughts and time, Mike

flag
4 
1) I don't see why you choose to single out Kurt Harris when there are many paleo-style, blogging MD's & PhD's. 2) You seem to be implying Harris is offering a more restrictive form of Paleo for otherwise healthy people when in fact his PaNu approach is less restrictive than most standard paleo approaches. (At PaleoHacks a paleo diet is dogma; therefore I assume comparisons are relative to the standard paleo diet unless explicitly stated otherwise. If you just happen to be picking Harris as a generic paleo advocate... um, very poor choice you need to do considerably more background reading.) – PortlandAllan Feb 17 2011 at 6:15
1 
I think Kurt Harris was choosen because poster respects him. Also I think all 12 steps PaleoNU is quite restrictive. – Ikco Feb 17 2011 at 10:52
@PortlandtAllan: I am not singling out Dr. Harris for any attack. Although I'm aware of other MD bloggers like the Drs. Eades, I found Dr. Harris' to be a well-defined approach (which, in a sense, means it is more restrictive) so I felt more comfortable evaluating his work than others. In sum, his writing is good enough to warrant consideration & criticism. – mac389 Feb 17 2011 at 14:53
How can you say you're not singling out Dr. Harris when you are singling out Dr. Harris? "Does Dr. Harris reasoning suffer from Berkson's Paradox?" doesn't leave much room for confusion. – PortlandAllan Feb 19 2011 at 4:11
@Portland: I'm not criticizing personally. You cannot get a specific answer without a specific question. Dr. Harris explains himself well enough to make his version of paleo able to be criticized. That's more than the MD bloggers who IMO ride on the clout of an MD without making medically substantive statements. So, I'll poke around his and ask questions because it's more likely I'd learn something. – mac389 Feb 19 2011 at 12:10

11 Answers

13

Does my reasoning "suffer"?

Cures for sick people aren't necessarily the basis of preventative treatment. So, what he sees work for sick people need not translate in advice for those who want to maintain rather than recover health.

My medical experience is 23 years of seeing -literally with the aid of advanced imaging- the effects of diseases of civilization, and how ineffective most standard medical therapy is for them. The ubiquity of the DOC and the inefficacy of standard medical care is what got me started reading about nutritional approaches to health 4 years ago.

When have I ever, even once, suggested a standard medical therapy for healthy people to employ in preventing disease?

I am in fact the one health blogger who does the exact opposite. I just wrote an entire post emphasizing how we should not assume the identity of "sick" as default. It's called "therapy versus life".

Have you even read the blog? I am very surprised that of all the pill and supplement pushing writers out there, you would name me as inappropriately applying therapy for the sick for prevention. I've never done any such thing. I am the only MD blogger I know of who recommends little to no laboratory testing. I OWNED a CT scanner and yet never recommended routine calcium scoring.

To push the envelope, why is his experience helping to diagnose the sick relevant to preventative medicine like proper nutrition?

Are you seriously asking how experience in diagnosing is relevant to preventive medicine? Would reading the same huge number of articles mined from pubmed but without medical training or experience make me less "biased" by not understanding what I'm reading about?

To further expand on my training - like most subspecialty trained radiologists, I spent at least half of my career doing general radiology, where you are responsible for knowledge of and diagnosis of literally every disease that can show up on a CT scan MRI or x-ray in patients of every age. This is a pretty broad grounding in human disease, and I would argue no specialty in medicine sees more of it.

If you think these insights are irrelevant to prevention, I would ask "how do you know what to prevent if you don't understand well what exists in the first place"?

link|flag
1 
@Dr. Harris: Thanks for the reply. My point resonates with your posts': normal physiology isn't the absence of pathology so the latter informs the former to a limited extent. Seeing what doesn't work may sometimes tell you what not to do, but the mechanisms of health need not be the negation of the mechanisms of disease. In essence I am asking the metonymous you the physiologist's question: What can't disease tell us about health? Amusingly, I responded to Uncle Herniation's comments with something akin to your last three paragraphs. – mac389 Feb 18 2011 at 21:39
1 
@Dr. Harris: Note how Steven Brecher highlights a sentiment that your stance doesn't come largely from your medical practice. – mac389 Feb 18 2011 at 21:43
6

I think Dr Harris could be biased somewhat about the prevalence or perniciousness of metabolic disorders due to the people he deals with in his practice. If you see NAFLD all the time, it's easy to have that bias your perception of the general population.

At the same time, however, I don't think his recommendations have much to do treating sick people. He's stated that he's "not a weight loss guy" and is more interested in helping people not hurt themselves than he is in healing them (something which he doesn't see as necessarily possible in all cases). His recommendations are deliberately generalized and when you read his blog you realize that in most cases what he's trying to get people to do is avoid excess fructose, seed oil and wheat - very general sorts of actions. He's not trying to preach to the sick with specific strategies to overcome various problems.

His relevant experience comes not from his professional career, but from his claimed reading of studies and literature relating to nutrition. If it turns out that he's been making all this up and never read anything, then he's a charlatan. So far, though, there's no reason to believe that. He seems genuine enough, and his claims are broadly in agreement with what you might find on your own.

As for Paleo and treating disease, it clearly depends on the disease. Switching to paleo will not fix tuberculosis, but it might help if you have a metabolic disorder or IBS. And since one of the foundational points of paleo is that eating correctly allows your body to function correctly, then the treatment is the same as the maintenance if it works; you just eat paleo and you're set. I don't see many people out there advocating severe diets for "treating" diseases and then different diets for "maintenance"; usually, if the "maintenance" diet doesn't work on the front end, it's not going to work at all.

link|flag
@pfw: I didn't mean to imply that Paleo diets would do anything against infectious diseases except inasmuch as they support a healthy immune system. Why is an MD qualified to interpret biochemical literature? Having been through medical & graduate school, I'm struck by (1) how little the scientific training an MD receives informs the daily practice of medicine and (2) how scientifically inept most doctors without scientific training are. Just because someone has read widely and has a related degree doesn't mean that one should take his word without debate on an open forum. – mac389 Feb 16 2011 at 20:04
I'm not suggesting that his MD brings him any qualification for nutrition science, and he has commented in the past that his medical education was worthless when it came to nutrition, so I think he agrees with both of us. Nor am I suggesting that one should simply take his word as gospel nutritional truth. It seems like you're broadly asking, "Why should I trust doctors, and since Kurt Harris is a doctor, why should I trust him?" I would submit that Kurt Harris very deliberately avoids acting in any capacity as a doctor on the internet, and so judging him as a doctor is inappropriate. – pfw Feb 16 2011 at 20:26
3 
Judge him instead as an interested reader of nutrition science who makes blog posts about what his interpretation of that science, thus giving you an opportunity to judge his acumen. – pfw Feb 16 2011 at 20:29
@pfw: Good point. I feel that as matters get more complex one increasingly relies on the presenter more than material, or some other heuristic. Perhaps the answer is for me to stop writing grants on neuroscience and move to nutrition (after tenure, of course). – mac389 Feb 16 2011 at 20:43
Haha, well, we'd welcome another interested reader of the science I'm sure :) – pfw Feb 16 2011 at 21:35
show 5 more comments
4

Hi, Mike/mac389, it seems you are asking two questions, and then some other questions have spun out of those in the answers/comments. I'll talk about your two questions, and try to keep my responses distinct.

First, on your primary (and first) question, I'm surprised that no one has given what seems to be the obvious response: Dr. Harris might see a lot of people with metabolic syndrome and related issues, but we all see a lot of people with metabolic syndrome and related issues -- because it's everyone around us. So the fact that he's a doctor doesn't seem to change his perspective all that much. (If anything the more relevant selection bias / confirmation bias might be the one we all have as paleos, doctors and non-doctors alike: we all see metabolic syndrome everywhere since we're looking for it everywhere.)

Second, as a response to what I think is your second question, the question about the consequences that might come from a Berkson's bias, since preventative care may not be the same as reparative care: Sure, I think that pfw is right for the most part to say: if you're healthy eat paleo, and if you're sick eat paleo. But even within paleo/real-food approaches there are some differences. So, for example the last section of the Jaminets' book is a fascinating (although speculative) discussion of the different things you might do with your diet and lifestyle to address extracellular versus intracellular infection. Similarly you might think that a particular species of paleo diet might be best to deal with cancer (see some recent threads on PH). Or you might think that you need to change your diet radically to heal your digestive tract before returning to a more inclusive paleo way of eating.

Those are my answers to each of the questions.

An additional question might be how the two questions relate to each other. As a general statement we can say that there are selection biases (seeing an unhealthy population in your office) that lead to bad medical advice. But in most cases it seems like they don't. Like paleo dietary advice, for example. Although even there there are exceptions: like the ones I gave above. (Or perhaps that it would be bad to switch very suddenly to a high-fat paleo diet immediately after a heart attack, etc.)

link|flag
Best answer - for being the most direct to the original point, and thereby showing the value of the question without the contentious aspects or personally-specific examples. This is (I hope) the intellectually sound kind of analysis and discussion that was desired by the OP, and certainly the kind desired by this reader. "No" is a perfectly valid answer to the titular question, and if well reasoned and well justified, a much more valuable answer for readers of this forum and Dr. Harris' posts alike. – uep Apr 7 2011 at 22:21
4

From the OP

"In sum, I'd be interested in the community's thoughts on how Paleo approaches to treating disease differ from maintaining health. In particular, other than for parsimony's sake, why should we assume them to be the same, when evidence exists that they aren't?"

I believe there is a trend towards moderation in much of the paleo blogosphere. Many now recognize that, say, potatoes or even rice are fine for those with healthy metabolisms. There are several bloggers who (I'd say) aren't really concerned with "paleo" per se but more about diets informed by tradition (chris masterjohn, melissa mcewen, paul jamninet, stephen of whole health source (sp?), etc etc). Stephen from whole health source recently said he was fine with limited gluten, rather than strict gluten avoidance.

This trend of moderation seems to answer your question, at least partly. As bloggers have recovered their health, there is a move to diets that are simply about 'maintaining health'. there is some disagreement with the idea of a 'paleo solution' (referring to robb wolf's book here), i.e. a diet that should be given to celiacs, diabetics, and the overweight, which is then prescribed for everyone.

Personally, I think the final step of moving to a maintenance diet has been dropping the term 'paleo' from my vocabulary. I am more inspired now by traditional food practices and in simply limiting gluten. I see no problem with having a beer or two here and there, even if it contains gluten. The idea of a strictly limited 'paleo solution' or 'whole9 30 day paleo plan' is just unnecessary for me at this point.

link|flag
The elimination of grains is huge not only from a pure Paleo perspective but due to allergies. Allergies to wheat and gluten are much more common than realized. That is one (but not the only) reason people get healthier going Paleo. – Mark V May 11 2011 at 15:07
4

Huh? Harris is a neuroradiologist by trade. Neuroradiology "is a subspecialty of radiology focusing on the diagnosis and characterization of abnormalities of the central and peripheral nervous system, spine, and head and neck." I infer therefore that his views on nutrition, exercise, and other subjects discussed here are not primarily and perhaps not significantly derived from his medical practice experience.

link|flag
@Steve: Dr. Harris' medical practice may little inform his views. But his training certainly exposed him to germane scientific principles. All doctors undergo the same training in medical school. After graduating from medical school, all radiologists still apprentice as internal medicine doctors (interns) for one year before studying radiology for 3-4 yrs. After that, neuroradiology is a 1-2 yr fellowship. So, the bulk of his education (basic science & internal medicine) are quite useful. Also, radiologists need to know considerable medicine to know disease stigmata to look for. – mac389 Feb 17 2011 at 14:57
@Steve: Your response raises a good point though- when is it appropriate for an MD to identify himself as such when not practicing medicine. Dr. Harris doesn't dispense personalized medical advice online. But, since unlike the Drs. Eades, his practice may not inform his views, why does he tell that he is a doctor if not to motivate us to consider his views because of his past training? To push it, what would you think of his blog if there wasn't an MD moniker on it. – mac389 Feb 17 2011 at 15:00
@mac389: I don't see any issues of interest, and none related to the original question, in whether KGH identifies himself as an M.D. on his blog. – Steve Brecher Feb 18 2011 at 19:52
@Steve: When someone says they are a doctor, I assume that people give his views on 'medical stuff' more credence they otherwise might. It's related to the original question because this type of selection bias was originally described in doctors who's patients motivated their research. Sounds familiar, no? – mac389 Feb 19 2011 at 12:13
2

A dietary problem (the SAD) is best addressed with a restoration of a diet that is consistent with human evolutionary design. Though we may think of food as medicine, it's not that the fatty steak is a delicious magic pill, it's that it is standing in for what was a highly toxic meal, such as a plate of spaghetti. We aren't feeding ourselves medicine, we are removing toxins and feeding ourselves food.

link|flag
@Travis: You make a good point that removing toxins can be healthy. But, there is a problem with thinking we 'evolved for' or that evolution gives the optimal solution. – mac389 Feb 16 2011 at 19:58
2 
I agree; I'll be the first to disagree with being overly presumptuous about what "we" evolved to eat, especially given the vastly wide array of possible environments and foraging strategies. That being said, it's safe to assume that precious little about corn-fed beef + high gluten wheat is advantageous for human health or consistent with our design. – Travis Culp Feb 16 2011 at 20:21
And it's not unreasonable to suggest that we did NOT evolve to eat a plate of spaghetti. – middlinplus Mar 1 2011 at 15:40
2

i don't think kurt harris has asked for his positions not to be debated. in fact, i think he's asked for quite the opposite. he has said that he gives what he considers conservative recommendations and tries not to get too far ahead of the research. i have no degree in any of this stuff but i fail to see where kurt harris' drastically differs from robb wolf, chris kresser, mark sisson or any of the other paleo bloggers out there. so, why are we singling dr harris out?

link|flag
@luckybastard: I'm not singling Dr. Harris out. To my knowledge, he is the only practicing clinician of those you mentioned. Whenever one lists the MD, there is a standard of conduct even if not duty to care. Is there a forum where one can debate directly with him? He does occasionally answer reader's comments. But, one can choose comments artfully. In truth I'm as underwhelmed by his research proof as by many of the SAD advocates. He cites no mechanistic studies nor trials that wouldn't reasonably suffer from a selection bias. So, is he speculating with big words? – mac389 Feb 16 2011 at 20:21
go check out robb wolf's site where they announced his participation in an upcoming podcast. he answered many questions there from laymen and other med professionals alike. i'm sure after his podcast airs, he will do the same. like most people who read dr harris, as a layman, who reads him and several other paleo bloggers and have tested out on myself the things they recommend. they seem to work very well. as far as other drs, chris kresser comes to mind as well as dr mark hyman who writes on huffpo(though not officially outed as paleo, his writings def line up pretty damn well). – luckybastard Feb 16 2011 at 20:50
@luckybastard: I agree that because scientific understanding of nutrition is laughable one must rely on empirical things like clinical studies and personal trials. But, from a debate stand point, perhaps you will admit that answering selected questions on a podcast is not the same as presenting a coherent thought to a critical audience for immediate unfettered debate. – mac389 Feb 19 2011 at 12:15
1 
"Whenever one lists the MD, there is a standard of conduct even if not duty to care." When I blog, I am a blogger, not anyone's personal physician. BLogging is not the practice of medicine. There is absolutely no standard of care or "conduct" for my bloggin any more than there would be for you to blog on any given topic. At best, blogging is just journalism. – Kurt G Harris MD Feb 22 2011 at 23:05
2

Paleo is a very young and emerging field. One of the main confusions is the difference between paleo for sick people trying to get well, and paleo for healthy people trying to get healthier (or perhaps trying to remain that way). Paleo interventions in either case are going to be different, but we don't yet have a way of making the distinction. I think your question reflects this confusion.

link|flag
1

If you are talking about degenerative diseases than it is really no different. Drugs usually treat symptoms of these degenerative diseases so drugs are not really a treatment.

If you are talking about some kind of infection such as cholera, I'll be the first one to ask for antibiotics.

link|flag
@aaa: I apologize for not being more specific. The most germane diseases are those like metabolic syndrome, diabetes, inflammatory & autoimmune disorders. Dr. Harris, for example, talks a lot about vascular health as indicated by calcium scores. It is important to realize that just as there are some people who will become diabetic no matter what, there are some people who will never become diabetic despite themselves. Why should what works for treating diabetes (or palliating its symptoms) work to prevent it except inasmuch as the underlying pathology is the same? – mac389 Feb 16 2011 at 19:53
What people would become diabetic no matter what? Type 1s? Even that's debatable. It would have weeded itself out of the gene pool, since it most often kicks in before reproductive age and we didn't have insulin to treat it. – Dana Feb 17 2011 at 0:11
@Dana: Your argument improperly affirms the consequent. There are type I diabetics in the world, so the trait hasn't not been weeded out. Alternatively, perhaps the trait was eliminated and then re-appeared. That suggests that it is advantageous or inexorably linked with some advantageous trait. Or, perhaps the incidence was decreasing until insulin therapy began, in which case, the selection pressure was quite low which seems at odds with the severity of the disease. – mac389 Feb 17 2011 at 15:05
@Dana: To clarify the "become diabetic no matter what" comment, I meant would become type II diabetic. However, because I can only cite those who progressed to frank/brittle diabetes despite following modern medical treatments some of which may exacerbate rather than eliminate the underlying pathology, I'll withdraw that criticism until someone shows that there are people who get type II diabetes or metabolic syndrome or experience no relief from it despite adopting a Paleo lifestyle. – mac389 Feb 17 2011 at 15:09
1 
LADA. A form of T1. Presents well into adulthood – Egk Feb 18 2011 at 16:11
1

It's not about strictly believing one person's opinion, no matter how scientifically informed. It's about looking for converging evidence from multiple sources. These can be bloggers, scientists, MDs, evolutionary theory, your own reading of the literature, etc.

link|flag
@Uncle Herniation: I agree that slavish belief is not useful. But, there is also a problem with cobbling together multiple sources. Sometimes they aren't comparable. People defer to experts because no one has the breadth of knowledge require to evaluate every piece of evidence independently. But, how do you know which experts to trust? And, how independent are your sources? – mac389 Feb 16 2011 at 20:24
1 
Critical thinking skills – Uncle Herniation Feb 16 2011 at 20:44
@Uncle Herniation: What do you mean? Critical thinking can detect logical inconsistencies and perhaps outlandish assumptions. Even Socrates noted how blind we are to our own biases & inconsistencies. (Perhaps that's partly why he detested books so.) How can someone critically evaluate something in which they don't have the proper training? – mac389 Feb 16 2011 at 20:56
The ability to find contradictions, just like the ability to synthesize well, is a lot rarer and a lot more effective than you think. There are a zillion things that are agreed on in physiology, but yet no one seems to be able to put it all together. Why is that? My friend in medical school told me he had two principal lecturers for their little unit on nutrition (just this year). One of them said: from all this you can see that the human body is meant to run on glucose. And the other one said: from all this you can see that the human body is meant to run on fat. Hmm ... Why do they disagree? – Paul Feb 16 2011 at 21:45
3 
You don't need to be an expert (or have proper training) in something to think critically about a subject. If you want to learn about something, read as much as you can and spend time thinking about it. – Uncle Herniation Feb 16 2011 at 23:22
show 5 more comments
1

Is there a known nutritional disease in which a type of eating cures disease but also causes it? I could see maybe a method of eating would cure one disease but maybe cause another. Touche on that one. Of course, if a lot of people switched to such a 'cure' It think you would then noticed a lot of this new prob taking it's place. Ane even if that were a danger, I think I'd still be interested in trying a cure for a disease that I might already have, even if it meant I MIGHT (but might not) get another one in the future. Because a disease I have now is probably more pressing than the chance of disease in the future.

Then again, paleo does not rest on just the theory that treatments might also make for good maintainance tactics, if if that is more probable than not. It also rests on understanding what we were designed to eat and always ate until recently, and it rests on what more healthful groups of people eat. And for everyone who has a problem, it also of course rests on treatment of those probs.

But my point is, paleo does not, or at least should not, rest on the ideas and assumptions of one person and one person's goals and concerns. Paleo is a process of taking in the whole milieu of data and considering it as fairly as possible. That may not be what always happens, but it's still a nice goal and one I like to think we do strive for.

link|flag
@Eva: I confess I disagree with the "designed to eat" sentiment because, by my understanding, evolution doesn't design for something so much as aimlessly tinker. There is no evolution god/demon asking how some evolutionary jump will increase some entity's fitness. Perhaps, then, in some sense, all foods are just tolerated? Evolutionary theory really plays with my conceptions of causality. – mac389 Feb 17 2011 at 15:19
2 
I think that Eva of all people knows that there is not actually a designer at work in evolution. The concept of design is just a shorthand for the non-purposive process of natural selection. You might then ask further to what extent we really are "suited" to a particular diet, but that is a separate question. – Paul Feb 18 2011 at 0:20
It is the difference between (say) 500,000 vs 5,000 years of exposure (with whatever selective adaptation that brings). Just as in all that time we haven't developed much of a tolerance for (say) the more poisonous funghi, some tolerance for (say) gluten has developed in the shorter time. Paleo and the modern research it draws on seems to be telling us that this is more unstable and not as widely and evenly distributed as has generally been assumed. The "designed" shorthand has been convenient until recently, but unfortunately is now too provocative for theologically neutral discussion. – uep Apr 7 2011 at 22:33
The best formulation of the practical use of this distinction was expressed by Matt Lalonde in a recent appearance on Robb Wolf's podcast. Incidentally, this was the first I listened to and my introduction to both proponents. Lalonde took the position that historical dietary analysis shouldn't drive choices directly, but can be very useful to form hypotheses about what we might best tolerate or find optimal, which can and should then be tested and investigated and understood with scientific rigour. Understanding the history of diet and eater can inform the design of a modern optimal diet. – uep Apr 7 2011 at 22:43

Your Answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.