How to use human history as a nutritional prior?

by CasioTheSane1 min read10th Mar 201250 comments

8

NutritionWorld Optimization
Personal Blog

Nutrition is a case where we have to try to make the best possible use of the data we have no matter how terrible, because we have to eat something now to sustain us while we plan and conduct more experiments.

I want to apply Bayes theorem to make rational health decisions from relatively weak data. I am generally wondering how one can synthesize historical human experiences with incomplete scientific data, in order to make risk-adverse and healthy decisions about human nutrition given limited research.

Example question/hypothesis: Does gluten cause health problems (ie exhibit chronic toxicity) in non-coeliac humans? Is there enough evidence to suggest that avoiding gluten might be a prudent risk-adverse decision for non-coeliacs?

We have some (mostly in vitro) scientific data suggesting that gluten may cause health problems in non-coeliac humans (such as these articles http://evolvify.com/the-case-against-gluten-medical-journal-references/). Let's say for the sake of arguing, that I can somehow convert these studies into a non-unity likelihood ratio for gluten toxicity in humans (although suggestions are welcome here too).

However, we also have prior information that a population of humans has been consuming gluten containing foods for at least 10,000 years, without any blatantly obvious toxic effects. Is there some way to convert this observation (and observations like this) into a prior probability distribution?

 

50 comments, sorted by Highlighting new comments since Today at 6:13 PM
New Comment

The Romans used cups made of lead, not knowing the effects of lead poisoning. It also took a while to notice the effects of cigarette smoking. If wheat gluten was Bad, but the effects weren't immediately obvious, it could easily go unnoticed for a long, long time.

(As far as I can tell, you can make at least a halfway decent case against basically any kind of food; it wouldn't be much of an exaggeration to say that science shows that eating, in general, is bad for you.)

It also took a while to notice the effects of cigarette smoking.

While tobacco was used by Europeans in some quantity since the early colonization of North America, it was the mass-production of cigarettes in the late 19th century that made heavy smoking a possibility for a large number of people. The first medical work showing a link between smoking and cancer came out in 1912 and 1929.

But smoking didn't get really huge in the U.S. until the mid-20th century, the era of mass media marketing of cigarettes (L.S./M.F.T.!); the British Doctors Study in 1956 was statistically pretty darned conclusive from a Bayesian standpoint, but was famously opposed by R. A. Fisher. Much of the later controversy on the subject of tobacco and cancer was manufactured by the tobacco industry, well after the matter was scientifically settled.

Sources:
http://en.wikipedia.org/wiki/Health_effects_of_tobacco_smoking
http://www.cdc.gov/tobacco/data_statistics/tables/economics/consumption/
http://en.wikipedia.org/wiki/British_Doctors_Study

Thanks for the update.

All foods both exhibit toxicity, and provide nutrients. I think the goal of nutrition is to choose the right foods in the right amounts to balance nutritional needs with toxic effects.

Unfortunately, this is easier said than done. When you really look into it we don't know much about human nutrition, but some choices are still better than others with the information we do have. The problem is I am having trouble finding a rigorous way to weigh these different choices. I guess the real question isn't "is gluten toxic?" but "are gluten containing foods more or less toxic than other alternatives which meet the same nutritional requirements?"

That's a great point about the lead cups, and smoking. It certainly makes me wonder what other things are hurting our health right now, that we potentially have the clues to identify but haven't managed to connect the dots yet.

There are other examples besides lead cups. Most nutritional deficiencies have only recently been recognized; scurvy's solution was famously lost for a century, and while there were folk remedies even back in the Roman era for curing goiters with seaweed, none of them indicate any understanding of the subtler effects on intelligence.

[-][anonymous]7y 0

"are gluten containing foods more or less toxic than other alternatives which meet the same nutritional requirements?

Nutritional and economic requirements. I'd guess gluten-free foods are no cheaper than otherwise-equivalent regular foods, so unless you're willing to spend an arbitrarily large amount of money on food that's also relevant.

I've often wondered if a large-userbase data collecting website could help solve problems like this by looking for very weak statistical correlations among coinciding events over large datafields. I.e. see how often people self-report eating X, see how often people self-report feeling Y, see how often one precedes the other and when they happen independently. The function to users would be letting them track their own actions (e.g. diet, health, etc) according to preset (or high-karma member-submitted) input:data -sets. I should think with members in the thousands such a thing would become useful. Especially were the service entangled with some social app to get users and some very good statistics processing to get results. Does anything like this at all exist? (Any obvious ideas why it doesn't, barring there possibly being lack of incentive to use it, lack of an incentive for a company/person to program it?)

Yes, it exists: http://genomera.com

They're actively running experiments and collecting data but are in "beta testing" and are very exclusive on whom they allow to join. I'm disappointed they didn't choose me when I filled out their request for a beta invite.

A huge problem with collecting data like this in the US population, is that everyone has a similar diet. There's so few people totally excluding gluten, you can't expect to measure it's effects with epidemiological diet surveys: you need to actually do a controlled trial where you tell people to avoid it.

In China where only about half of people eat foods with gluten the biggest epidemiological study ever performed (the China Study) did find that wheat intake was independently correlated with overall mortality (http://rawfoodsos.com/the-china-study/). They never published this finding themselves, but the correlation is clearly there in the data.

There's a lot of question about their methodology- they didn't keep or report data on individuals, but lumped whole communities together as single data points. There's likely a lot of highly correlated regional habits that weren't on the questionnaire, and I tend to find the whole study pretty questionable. For the most part, it's just comparing the health of rural farmers with wealthier urban Chinese- the two groups have radically different health, lifestyle, and diets and we can only control for the few questions they actually asked.

Perhaps now that gluten avoidance seems to be becoming a "fad diet" in western countries, suddenly it will be possible to actually collect good data on this.

That looks like it could prove really useful / interesting; thanks for linking.

I guess the entry requirements for beta are strict because they're trying to keep to a small set of variables for the people to check? It would have been really interesting to spy in on though. Regarding the China study, it sounds either like there was no effort to control for other obvious/statistically-true correlates or that there is no possible overlap at all to abstract a controlled comparison from. A fraction of that data might be useful (all data is useful! ...yum!). I think with sufficient (though perhaps improbably large) sample size even user-submitted data with large amounts of noise becomes useful. Any empirical paradigm more open and faster than the current is bound to be a good thing, even despite inaccuracy, for reasons of sheer brute force.

At least with user submitted noisy data you have individual data points, and potential to track individuals over time... unlike the China Study where entire communities were just averaged into a single point.

There's some usable information in the China Study, but not as much as people think... it's being touted as "proof" that all animal-based foods cause cancer (in a popular diet book by the primary investigator Dr. Campbell) because the two were well correlated in the data, when it's nothing of the sort.

The folk wisdom of "eat a bit of everything" seems to go well with humans being omnivores and with not overdosing on any particular harmful ingredient. That was before pollution and allergies were rampant, of course.

The folk wisdom of "eat a bit of everything" seems to go well with humans being omnivores and with not overdosing on any particular harmful ingredient. That was before pollution and allergies were rampant, of course.

This would seem to require that most of the 'wisdom' being embedded in the 'everything' in question.

Good point, foods that have been historically clearly identified as toxic would have already moved into the non-food category. So this advice is effectively saying, "look only at your priors and don't consider or collect new data."

If our ancestors took this advice literally they'd have kept eating things that can obviously kill you, like digitalis.

In reality, I think the spirit of the advice is a warning not to under-estimate the importance of your priors: don't consider new data in isolation.

In reality, I think the spirit of the advice is a warning not to under-estimate the importance of your priors: don't consider new data in isolation.

You have managed to extract a rather useful generalizable point from the subject!

As well as this we can assume that the advice also covers the idea of not putting all your eggs (be they epistemic or dietary) in one basket.

That only makes sense in cases where you have no additional data suggesting that some foods are healthier than others.

Also, if a few foods exist which exhibit chronic toxicity at low doses, but you don't know which foods those are, wouldn't it be safer to limit the total number of different foods you consume, as to limit the chance of consuming a particularly bad one by chance? While atypical, there could be cases where the toxicity curve is relatively flat, and lower doses don't really protect you. For example, an endocrine toxin that simulates a hormonal signal even at a low dose.

The thing about toxicity though - not all poisons have any effect whatsoever in low doses. Some do have effects that don't go to zero at low doses - radiation for example - because it does not dilute below 1 particle track through nucleus, which is harmful. The carcinogens in general, as they don't dilute below causing 1 mutation to dna in 1 cell, which has probability of harm. But organic compounds which are not carcinogenic - not as in natural but as in organic chemistry - and especially those from plants, they may well be entirely harmless at the doses below those that actually result in toxicity. Of course there may be situations where they are 'the last straw', but in general not everything is harmful all the way down.

It's not quite that the toxicity effects don't go to zero at low doses for things like radiation — the effects of low radiation exposure may indeed be zero. The toxicity may be zero at even relatively high doses. Rather, the value of radiation's dosage curve at the low end is stochastic, where for chemical toxins the value of the dosage curve on the entire domain is the severity of toxicity.

Radiation actually appears to exhibit hormesis- small doses of ionizing radiation likely activate some protective response, which decreases rather than increases cancer risk.

I think allergens, and compounds which initiate autoimmune disease (gluten in coeliacs) are a good example of something which remains toxic at low doses. There hasn't been a lot of research into it, but even extremely low doses of gluten seem to be harmful to coeliacs.

Well, that's other way to see it. The effects of carcinogens at low doses are increase in cancer probability, and is almost definitely linear (as it is added to existing mutations, and you are effectively dealing with a small piece of a mutation vs cancer risk curve, at a non-special point).

You have to be careful what you mean by carcinogen. I suspect what you're saying is true for some chemical carcinogens and ionizing radiation in single burst exposures, but not for other time varying patterns of ionizing radiation exposure.

It hasn't been confirmed in humans, but ionizing radiation exposure most certainly doesn't have a linear dose response with in vitro cell lines. Well, it does for single doses- but initial doses to be protective against future doses (hormesis): http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1533272/?tool=pmcentrez

1: that happens for quite significant doses only (1 centigray = 0.01 gray = 10 millisievert for gamma and x-rays = 1000 hours of exposure to 10 microsievert/hour background)

2: on single cell level there is no small doses, there's small probabilities of 1 particle track through nucleus (edit: or 1 Compton scattering electron track for gamma rays). Non-linear response to probability would indeed be quite extraordinary.

Ionization of DNA molecules is going to be linear in response to radiation, but you can't assume that will result in a linear cancer risk. Cells can up and down regulate their DNA repair mechanisms under varying conditions, allowing for a much more complicated relationship between radiation and cancer risk.

In this case, cells appear to up-regulate poly (ADP-ribose) polymerase (PARP) in response to past radiation exposure, increasing the chance of an accurate repair after an ionization event damages one strand.

//edit: The nucleus is also not the only potential site of cancer inducing mutations, the mitochondrial genome is another possibility.

Before cell can regulate it's DNA repair due to radiation, it has to detect the radiation, which at low levels in question (say, up to ~10x natural background, that's ~ 1..5 microsievert/hour, 2 000 .. 10 000 times less than centigray of gamma per hour) means detecting the probability before anything happens.

The issue with radiation is that people don't understand the units. You read the study you linked, you see, 1 centigray of x-rays, that's a 'low dose' they say, in god knows what context (Radiation therapy? Sure it's a low dose there). That's a 10 milliSieverts, okay? The average background dose a human receives per year, is http://en.wikipedia.org/wiki/Background_radiation , or 1/4 of that. Nobody's been proposing that 4 years worth of normal dose in a hour are going to still be linear.

edit: or actually, we do. We interpolate the low dose effects from the doses of somewhere around 0.1 Sv and up, based on various real world human data. Meaning that, if the effect outlined in your link is real, and there are some defence mechanisms activating at 0.01 Sv which prevent some of the DNA damage (at some other expense) - then we are underestimating the carcinogenicity of radiation at the low (near background) level, at which those defence mechanisms are not active. That is kind of scary to think about, in terms of potential extra cancer deaths.

Okay, I was misunderstanding what you were saying, and it makes sense now.

To paraphrase: Cancer risk in response to radiation levels can only be non-linear when the cell sees past radiation damage signaling it to mount a response. At low doses a given cell is unlikely to see any DNA ionization events, and therefore the risk must be linear.

That's a great point about the potential problem with extrapolating low doses from high dose data. That should really be investigated more carefully... if true "minor" radiation exposures could be a lot more risky than existing estimates suggest.

Yes. Instead what is happening, the nuclear lobby is citing studies like the one you linked as evidence of "radiation hormesis" i.e. generally beneficial effects of "low doses" of radiation (in your study the cells and the controls have to be irradiated with high dose afterwards to show any benefit; i'm very dubious that there exist any benefit versus the background), and lobbies for removal of strict EPA limits.

It gets even worse than this; the definition of Sievert and the procedures for calculating exposures of people rely on linear model - if the effects are linear then the average dose is all you need - and what the lobby wants is to use linear-model justified average doses together with threshold model, which just doesn't make sense. There literally can not be any consistent non-linear response to doses in Sieverts because the concept of dose and concept of Sievert presume linearity when it does not matter how the dose is distributed in space and time (well, up to organ scaling factors).

I've certainly seen several politicians argue that "radiation is actually good for you," but I've yet to hear any actual radiation health physicists argue that point...

Well, one doesn't usually see any actual radiation health physicists argue anything. I sure seen various engineering type people argue its good, and there are entire countries (Japan) where the linear-no-threshold model is evidently not adhered to.

Plus there is something weird going on with wikipedia articles on the subject all trying to present the pre-LNT views as something new that's challenging the LNT, complete with editing out of highly relevant historical references. Then there is "radiation hormesis", a hypothesis, that the radiation is good for you. Not "because of such and such specific response, radiation is good for you" - just a hypothesis that it is (which incidentally is the first "hypothesis" that comes up when a new exotic poison is found: someone hypothesises it to sell it in small amounts as a cure). Except that its presented as something new. Complete with a laundry list of rationalizations of how it might be so. That's terrible, and misleads people a fair lot.

I dunno if I should go ahead and write article on the topic.

Then there is "radiation hormesis", a hypothesis, that the radiation is good for you. Not "because of such and such specific response, radiation is good for you" - just a hypothesis that it is

Is this a fair description of the history and science behind hormesis?

See for yourself:

Is it more like some new outcome like 'ohh, there's new method by which the cell would know the radiation doses at low near background level, even for alpha particles a single of which does giant damage! Some new exciting physics discovered - the quantum probability can be measured before event happens!. That got to be useful for something, maybe for defence response. Ohh, there is the defence response, and its so strong.... I wonder if low doses of radiation are good for you?'

Or is it more like like 'okay, suppose the radiation is good for you, let's think and come up with justifications, okay, the untapped powers of organism that will be'.

The former is the process of scientific enquiry, the latter is the process of pseudoscience - start with desired effect, make up vague cause, later on perhaps think up a zillion specific causes, good luck proving them all wrong. I wonder why we even take obvious products of entirely backwards reasoning at face value as if they were not fundamentally different from products of forward reasoning?

Also try calculate how many people are required to find LNT-predicted dose effects at 10x the background. There aren't going to be direct evidence. There will be very strong indirect evidence, such as difficulty for the cell to measure doses near background, and generally low prior probability for some magical untapped powers of organism.

I have seen for myself. And I've also read about all the other hormesis effects like cold, caloric restriction intermittent fasting, exercise etc, which make a potential hormetic effect from radiation quite plausible (regardless of your sarcasm about 'magical untapped powers').

I wonder why we even take obvious products of entirely backwards reasoning at face value as if they were not fundamentally different from products of forward reasoning?

I don't think it's backwards at all.

You're a very dismissive person, I think. I give you a link with all sorts of modern results and mechanisms, and all you do is speculate wildly about ulterior motives.

I have seen for myself. And I've also read about all the other hormesis effects like cold, caloric restriction intermittent fasting, exercise etc, which make a potential hormetic effect from radiation quite plausible (regardless of your sarcasm about 'magical untapped powers').

It is entirely implausible that there would be a potential hormetic effect from 'radiation', because there isn't single such thing.

There's the beta radiation. There's the gamma radiation that mostly knocks off electrons (works a lot like beta, but wavelength-dependent in how the damage is delivered at cellular scale). There's the alpha radiation, that is massive, stops very quickly, and generates immense damage in a short path. That's like a hypothesis that 'fruits' are good for you, right here. Under the linear no threshold, you can bin all radiation types into one box (after you apply scaling factors). Under hormesis you can't.

You're a very dismissive person, I think. I give you a link with all sorts of modern results and mechanisms, and all you do is speculate wildly about ulterior motives.

You give me a link to wikipedia article with a non labelled 'graph' near top. Is it logarithmic scale? Is it linear scale? What are proposed dose values, even approximately? What is the proposed type of radiation, is it all types, is it one type, what energy range? [note that those are only abstracted out, after applying scaling factors, under linear model]

A hypothesis follows from a theory regarding the mechanism, not other way around. When different theories generate same conclusion, that's still different hypotheses, because the explanations differ. When an observation is made, it's called fact. A fact can generate multiple theories, that may generate other testable hypotheses, that would allow to discern between theories.

When a hypothesis is generated from the argument about all the other hormesis effects (never mind all the other non-hormesis effects!), which is the only explanation as of why this hypothesis would be so broad as to include all radiation types, and then multiple independent ostensibly 'mechanisms' aka theories are generated, that's called pseudoscience. Especially if unlabelled graphs are produced. A hypothesis has to be testable, i.e. possible to be shown wrong, that means the hypothesis has to specify some particular dose range, where if it is not found, the hypothesis (along with theory that produced this hypothesis) is put to rest as false. And note, the pseudoscience is not just cranks who are obviously wrong. They are non-science. Pseudo-science is something that looks like science, to anyone who lacks domain knowledge, and often, even to those who do have domain knowledge.

The key that distinguishes scientific propositions, from pseudo-scientific ones, is testability. The hormesis is pseudo-testable - you have to search over giant combinatorial space of 3 radiation types, different energies, and many orders of magnitude of doses, before you can falsify it.

It is entirely implausible that there would be a potential hormetic effect from 'radiation',

It's perfectly plausible; why are you making some assumption like 'all forms of radiation in all tissues have the same hormetic effects?' Such strawmanning does you no favors.

because there isn't single such thing.

...Which the article prominently discusses as part of the French review. Not trying very hard, are you.

Under the linear no threshold, you can bin all radiation types into one box (after you apply scaling factors). Under hormesis you can't.

So it's just as much a point against the simple linear model as it is against hormesis.

You give me a link to wikipedia article with a non labelled 'graph' near top. Is it logarithmic scale? Is it linear scale? What are proposed dose values, even approximately?

That graph is explicitly labeled as hypothetical and a visualization! Holy cow. Do you never visit Wikipedia? They use diagrams all the time to explain stuff (and not as primary source material & proof...)

You want numbers for particular setups with particular organisms, that's what the studies section is for!

/ignores naive philosophy of science and name-calling

It's perfectly plausible; why are you making some assumption like 'all forms of radiation in all tissues have the same hormetic effects?' Such strawmanning does you no favors.

The hypothesis goes as following: fruits are good for you. Find 1 always-poisonous fruit, that disproves the hypothesis. That's how science disposed of this particular form of mindless arguing over what is, and isn't, a strawman. Now, 'there is at least one kind of fruit, that is good for you, in at least one specific amount', that's entirely different hypothesis, and must be stated as such.

edit: but that hypothesis should leaves you wondering, how do you know that there is at least one, rather than some specific one.

With regards to the listed cellular repair mechanisms, they are each a separate scientific hypothesis, testable one.

edit: also, for your information. I've added some of the cautious-sounding parts of the article myself a while back. So thanks for compliment.

anyhow, i guess i came off wrong there. Let me explain. The issue here is the technical one. The 'hormesis' is wannabe policy change motivated proposition. The specific cellular response mechanisms and such, are scientific hypotheses, plural, testable, and none of scientists who are genuinely studying radio-biology are eager to reduce his study to a hypothesis that repair overcomes the direct damage by radiation, or that the whole mechanism results in the net benefit for the organism. That is important distinction here. Hormesis is complete pseudoscientific BS. Those things, are genuine hypotheses, and make testable predictions that at a specific dose, of specific type of radiation, there would be some particular mechanism activating, and that strength would be such and side effects such and such. Eventually it can be found in which conditions we should expect net benefit, and what side effects there may be (e.g. immune system stimulated, less cancer, more auto-immune disorders). Those folks aren't eager to show most 'benefits' and hypothesise in the benefit direction. Their theories may be most directly testable via negative results other than cancer. The Hormesis cherrypicks out of those studies some net benefit. It may still turn out, that we will be willing to trade the negative effects, if any, for reduced cancer rate, if ever - or not. That is up to real science to decide. Not up to 'hormesis' studies.

Now, 'there is at least one kind of fruit, that is good for you, in at least one specific amount', that's entirely different hypothesis, and must be stated as such.

Indeed it is. Since you are aware of this, one wonders why you choose to uncharitably interpret hormesis in the broadest simplest possible way.

edit: also, for your information. I've added some of the cautious-sounding parts of the article myself a while back. So thanks for compliment.

So your disingenuousness is all the more irritating and willful.

Their theories may be most directly testable via negative results other than cancer. The Hormesis cherrypicks out of those studies some net benefit. It may still turn out, that we will be willing to trade the negative effects, if any, for reduced cancer rate, if ever - or not. That is up to real science to decide. Not up to 'hormesis' studies.

So to rewrite what you are saying, "Hormesis - as anyone but me defines it - exists as an effect in studies, but it's still all pseudoscience and 'real science' needs to investigate it."

Go read "biological effects of ionizing radiation" or some mainstream knowledge on the topic, please, before asserting that my opinion is not mainstream, and what is the mainstream view of hormesis is, and rewriting what I am saying.

Especially, do some homework assignments so you have some idea of e.g. how often a cell is hit by 'radiation' at near background level, so you see how plausible it is that the low doses - which affect things stochastically - have any sort of nonlinear effect.

Then delve into controversial stuff like hormesis.

That's actually a great heuristic: do not learn science from controversial stuff. Learn topic properly first. Before learning radio-biology, learn some physics, and biology. I'm pretty sure I seen an article to this sense on lesswrong. The reports like BEIR, that's the summarization of knowledge by experts. Wikipedia is a lack-of-scholarship bias. The wikipedia article on the topic would of been best balanced and most accurate by citing precisely 1 mainstream meta-report. Any balancing has already been done by far more qualified experts than Wikipedia editors.

Go read "biological effects of ionizing radiation" or some mainstream knowledge on the topic, please, before asserting that my opinion is not mainstream, and what is the mainstream view of hormesis is, and rewriting what I am saying.

The existence of hormesis in some examples to some degree is not disputed in the mainstream. All the arguments are about how widespread it is, how important it is, what the mechanism is, whether particular studies properly controlled for various issues etc. In other words, 'normal science' in the Kuhnian sense.

If you want to claim something like 'the mainstream view of hormesis is that no form of hormesis exists to any degree in any context whatsoever' and that hormesis is entirely outside the current paradigm, I am going to have to call BS on that and demand citations.

The existence of hormesis in some examples to some degree is not disputed in the mainstream.

False.

http://www.nap.edu/openbook.php?record_id=11340&page=332

and read rest of it.

The first thing:

Hormesis has been defined as “the stimulating effect of small doses of substances which in larger doses are inhibitory.” As stated by Wolff (1989) the meaning has been modified in recent times to refer not only to a stimulatory effect but also to a beneficial effect.

The wiki disposes entirely with stimulatory meaning. The stimulatory meaning is sensible. Stimulation may, or may not result in positive response (typically doesn't, can result in super linear negative responses just as well), which depends to type of stimulation. The beneficial one, is pseudoscience, and term hijacking as well, to list all the studies in the stimulatory sense as support.

BEIR V discussed the effect of confounders and the ecological fallacy2 in the evaluation of high-background-radiation areas and concluded that “these two problems alone are enough to make such studies essentially meaningless” (NRC 1990).

...

Ordinarily, epidemiologists do not consider ecologic data such as this as being sufficient for causal interpretations. Since the data are based on populations, no information is available on the exposure and disease status of individuals. Such data cannot be controlled adequately for confounding factors or for selection bias. Although ecologic data may be consistent with an inverse association between radiation and cancer, they may not be used to make causal inferences.

which is how you call something "pseudoscientific BS" when you are writing that sort of report.

And finally, after politely saying many very bad things about it,

The committee concludes that the assumption that any stimulatory hormetic effects from low doses of ionizing radiation will have a significant health benefit to humans that exceeds potential detrimental effects from the radiation exposure is unwarranted at this time.

Emphasis mine. Also, going to edit the wiki sometime with the latest BEIR, this replacement of stimulatory meaning with beneficial (without citing that it was originally stimulatory) on wiki is certainly not a good thing.

The whole notion of beneficial effect needs to go way down as a speculation that positive effects of multiple stimulations will outweight the negative effects, and the curve might look so and so, albeit the dose range is unknown.

And your cherrypicking continues. Your link specifies repeatedly that it is only discussing radiation hormesis, and drawing on sources about radiation hormesis. Radiation hormesis is the most controversial form of hormesis. Notice that they didn't conclude that no form of hormesis exists, which they really should have since a fortiori that subsumes any discussion of radiation hormesis...

which is how you call something "pseudoscientific BS" when you are writing that sort of report.

It makes perfectly sensible criticisms of epidemiology, which apply to all sorts of things like the War on Salt or cholesterol and not just radiation hormesis. I agree, such studies are extremely weak.

But do we just have epidemiological correlations for radiation hormesis - much less all the other forms of hormesis?

Ahh, okay, sorry, I misunderstood your point as applying to 'radiation hormesis', because that was what i was originally talking about.

Other hormesis - sure, some chemicals have hormesis effects. That's a fact. Much misused fact (by homoeopathy practitioners), but a fact. Implies absolutely nothing about radiation hormesis. edit: and i don't see why even have a mental bin 'hormesis' for vastly different mechanisms by which something bad may be good in small doses.

edit:

I agree, such studies are extremely weak.

I would say, they are essentially meaningless. Essentially is empty word that you add not to sound too nasty.

edit: by the way, on other hormesis. The original thing I has posted in this thread, was:

The thing about toxicity though - not all poisons have any effect whatsoever in low doses. Some do have effects that don't go to zero at low doses - radiation for example - because it does not dilute below 1 particle track through nucleus, which is harmful. The carcinogens in general, as they don't dilute below causing 1 mutation to dna in 1 cell, which has probability of harm. But organic compounds which are not carcinogenic - not as in natural but as in organic chemistry - and especially those from plants, they may well be entirely harmless at the doses below those that actually result in toxicity. Of course there may be situations where they are 'the last straw', but in general not everything is harmful all the way down.

I did specifically note that just because radiation and carcinogens have linear effect (as commonly accepted by mostly everyone in the field), does not mean other poisons do. You absolutely can't accuse me of being unfair towards other examples of 'hormesis', (albeit it puzzles me why one would bin all those drastically different mechanisms and effects together under same label)

Peace, now? I know my biological effects of radiation stuff. I am passionate about the subject because the failure (unwillingness) to apply LNT results in preventable human deaths. It is not abstract problem. I don't bring deaths up because there's accusations of emotional bias. Can i be allowed to be passionate about subject with deaths attached, if I am entirely (and very boringly) mainstream in my scientific opinion, AND actually have argumentation I understand? All the stuff of how hard for the cell it is to detect the 'low doses of radiation' (which are a track once in a blue moon). Other stuff like this. It's a complicated subject. The radiation been fascinating me since I was little, and I read on it as hobby. And i study from mainstream sources before reading controversial ones.

Ahh, okay, sorry, I misunderstood your point as applying to 'radiation hormesis', because that was what i was originally talking about. Other hormesis - sure, some chemicals have hormesis effects. That's a fact. Much misused fact (by homoeopathy practitioners), but a fact. Implies absolutely nothing about radiation hormesis

You dismissed the area of research as unscientific self-serving wishful-thinking pseudoscience, which is demonstrably incorrect and sits poorly with me as devaluing the meaning of terms like pseudoscience which should be reserved for pernicious things like homeopathy. Radiation hormesis is highly controversial, may not be true, and has methodological issues with some of its results, but this is all 'normal science' and not pseudoscience, and looking into it is perfectly scientific.

If nothing else, the other areas of hormesis mean that it is more plausible that radiation hormesis would also exist and looking for it may not be privileging the hypothesis. If no one had ever found any kind of hormesis, that makes radiation hormesis especially unlikely! and so by symmetry, finding other sorts of hormesis make radiation hormesis a little more likely.

You dismissed the area of research as unscientific self-serving wishful-thinking pseudoscience, which is demonstrably incorrect

I dismiss the health benefits notion. Did you read that BEIR link? How come they aren't convinced? Of course they stay polite and don't call anyone self serving. That's not their job to look into motives.

and sits poorly with me as devaluing the meaning of terms like pseudoscience which should be reserved for pernicious things like homeopathy.

Pseudo-science is something that looks like science to careless contemporary observers, okay? That's the original meaning of the word. Phrenology looked like science to contemporaries. Freudian psychoanalysis did. Bloodletting did. And so on. None of those topics were readily recognized as cranky by contemporary average person (even though the scientists generally dismissed them as unscientific, non-testable, and non-predictive)

Homeopathy barely looks like science to anyone any more. It's a non-science, and much of the followers cherish the notion that it is non-science (they say, we ain't trusting scientists).

I, instead, find that restriction of notion of pseudoscience only to cases plain obvious to most careless contemporary observers (us), devalues the notion. The pseudoscience is something that can fool you, or that can fool me. It can't fool most scientists, hence the BEIR very briefly dismisses all of it [rad. hormesis]. It can fool average person, hence wiki doesn't do the same.

If nothing else, the other areas of hormesis mean that it is more plausible that radiation hormesis would also exist and looking for it may not be privileging the hypothesis. If no one had ever found any kind of hormesis, that makes radiation hormesis especially unlikely! and so by symmetry, finding other sorts of hormesis make radiation hormesis a little more likely.

See the original post of mine that started this giant controversy. I explained why, exactly, one shouldn't do precisely this, the other way - conclude negative effect of low doses of poisons, from the negative effects of low doses of carcinogens. The physics here is not symmetrical.

I was going to continue this, but then I noticed that almost all my karma is gone (!) and most of my comments in this thread are at -2; your comments are also doing pretty poorly. (Not my doing. I try not to downvote any comment I am replying to.)

If I keep this up, I won't be able to post and my commenting will be rate-limited (which I hate), and the fact that we're both getting downvoted indicates LW in general finds this a worse-than-worthless discussion, so... I'm going to bow out, as much as I dislike not having the final word.

EDIT: and this gets upvoted almost immediately! You know guys, you could just leave a comment like 'I don't think this discussion is very productive, maybe you should stop' and not rely on the karma system with its crude numbers to convey your desires.

I agree with you on hormesis being plausible. But the two of you generated some fairly low-quality discussion. The downvotes are understandable.

People are giving you karma because they don't actually want you to suffer real harm (rate limiting).

Most people would steer clear of interfering in a street brawl. An anonymous downvote is like calling the cops :)

"real harm"?!?

Ok, that was funny. I just meant to indicate that it seemed a heartfelt and direct plea by Rhwawn, a person acting in good faith.

they are afraid of karma system, what if we both downvote them together. Also they are still reading, implying some sort of interest in the issue. My karma loss is mostly from posting something controversial in main for 7 downvotes, 70 points.

edit: btw if someone's coming in consistently to downvote, that person doesn't want discussion to stop as easy way is to stop coming in, it's not like this expands all the way. edit: could even be worse. downvoting so we think we are downvoting each other, hence get pissed off more.

[This comment is no longer endorsed by its author]Reply

edit: btw if someone's coming in consistently to downvote, that person doesn't want discussion to stop as easy way is to stop coming in,

This does not follow. People who prefer a conversation to stop most certainly can be expected to downvote comments in that conversation. It works.

(In case Rhwawn was wondering, this does not suggest that I am the one who downvoted them. I don't recall doing so in any near context.)