Friday, October 5, 2018

Thirteen days of near-carnivory fails to impress

Alan Levinovitz is an author and scholar of religious studies who has written about diet and science. He recently wrote an article criticising the carnivore diet: The Carnivore Diet Is the Latest Fad to Ignore That Food Does More Than Just Feed Us. It is this article that got me talking with Alan on Twitter about his suggestion in the article that the anecdotal effects are psychological. I was delighted that he took me up on a long form debate about this, and have posted that separately. I admire and applaud Alan for his cordial manner on social media, which is a precious rarity.

Here I want to object to a couple of points in his article that have to do with keto-adaptation, not carnivory in particular, and not the whole muddy question of potential disease modification. I absolutely agree with his cautions that we need more study, that proposed mechanisms are routinely falsified, and that the baseline Western diet combined with a wretched medical system incentivises junk science cures. However, I don't see how his 13-day trial sheds light on the biology, and I think he should have done just a little more research.

His article doesn't stand alone. It is consistent with prevalent and justified attitudes that nutritional science is mostly bunk. Unfortunately, this often results in a base level of hostility and refusal to even consider a scientific perspective on a topic viewed with suspicion. This is exemplified by a Twitter interaction I had in which someone tweeted Alan Levinovitz's article, saying that she was worried about the effects of ketosis on the brain. I was eager to share some of my research, because if there is anything the science shows, it's that ketosis is healthy for the brain! Fetuses and breastfed babies are in mild ketosis during brain growth, neurological diseases and cognitive decline are sometimes reversed by ketosis. It is our brains that probably account for the fact that we alone as species can stay in ketosis without caloric or protein restriction. When I invited her to read my presentation on this from earlier this year, she acted as if I were a salesperson approaching her on the street. "Not interested. Have a nice day."

https://2.bp.blogspot.com/-OTeaA7eeXV8/W3ehACCL0RI/AAAAAAAAGiY/UVeALdFCBu4aksRHCT-2tjkRezEqQq-qwCLcBGAs/s320/Dr-Jen.png
https://2.bp.blogspot.com/-1rYkyVlFPzQ/W3ehBjbfzmI/AAAAAAAAGic/1aYhFmid63oN14ZI1DHKdXorAdUhe5EkwCLcBGAs/s320/response-to-dr-jen.png

It turns out that ketogenic diets are drastically different from other diets, because they directly affect metabolism. As I've explained in one of my first posts on the topic, ketogenic diets are recognized as being at least as powerful as drugs in at least two major clinical condition, and there is reason to believe they are useful against others. In the noise of weak dietary therapies, it's harder to see a signal. This difference, though, doesn't happen all at once. Keto-adaptation accounts for the two phenomena Alan described that I want to address.

Since we're already on a tangent looking at that tweet, let's just talk about colon cancer and fibre for a moment. I won't delve into it deeply here. You can read my article about the misapplication of science about butyrate from plant fibre to colon health, if you're interested in that angle. Note also, though, that associations between fibre intake and colorectal cancers, which is confounded by the healthy user profile among other things, show minimally reduced relative risk—far above 0.33 [Mur2012] which is considered by some a minimal threshold for even suspecting causation (see e.g. [War2012]). It is understandable to "worry" about colon cancer, given the hype about it, but you would think that someone in the medical profession would want to dig a little deeper and not just parrot headlines.

https://3.bp.blogspot.com/-ISWkQVtNx0s/W3eg6spxcPI/AAAAAAAAGiQ/lvbz6oSAntgmFX8N1sTcIpEdZcA6AK49ACLcBGAs/s320/relative-risk-zones-of-interest.png

Bullshit pounds lost

Alan explains his experience of losing 7 pounds while carnivorous, but gaining back 5 after resuming his high carb diet, as follows: "Those were the bullshit pounds that low-carb diets use to hook their marks, like a con man giving away the first jackpot in a street hustle... If you stop eating low-carb, or “zero-carb,” the early loss—up to 12 pounds!—returns instantly because it’s not fat you’ve lost, just water released as your body burns through its glycogen stores."

Indeed, as I described in the second link above, much of the first few pounds lost when you stop stuffing your liver with carbs in the form of glycogen are from water required to store it. You lose carb weight before your fat loss will accelerate on a ketogenic diet. Your body disposes of excess carbs before excess fat just as it does in the bloodstream when your blood sugar is sufficiently high. However, that hardly impacts the fact that you lose excess fat on a ketogenic diet. People like me, who have lost over 60 pounds, or others who have lost even more, did not lose just water, and it would be ludicrous to suggest that. Total weight lost will only be mostly water if you stick to the diet for a bullshit timeframe, like 13 days. This isn't just a quick-fix, 13-day, fit-into-your-wedding-gown hack.

In any case, you could make the same argument about fat! Fat is stored with about equal parts water. When you lose a pound of fat, that's actually half water [Mel1942], twice as much as what carbs are stored with. What a scam, right? Granted, Alan is right that it's only carbs and the necessary water that pack back on so fast. Apparently he lost 2 fat pounds in 2 weeks not counting the glycogen-water, presumably eating as much as he liked, and he doesn't appear to be particularly overweight, so one wouldn't expect more true loss than that, especially in such a short time frame.

I think this criticism is only fair if you've been misinformed about the diet in the first place. Understanding the basic physiology of how the diet works would seem to me to be the onus of the dieter. Writers advocating the diet tend to clarify this. For example, in the most recent Atkins Diet book from Westman, Phinney, and Volek: "As with any weight loss program, some of the initial weight loss you'll experience is water weight... So if you drop 10-15 pounds in the first few weeks, you'll be saying good-bye to some unnecessary water weight along with the initial fat pounds. But once that excess water is gone, you'll be losing primarily body fat." From yours truly: "Another noticeable effect in the first days is water loss. One of the inefficiencies of glycogen storage is that it needs to be stored with water. It takes about 3 or 4 grams of water to store a gram of glycogen [ref] . This means that as you deplete your glycogen stores you could lose up to 2 kg of water! Not only that, but high circulating insulin levels cause water retention by inhibiting sodium excretion (see e.g. [ref]). The keto diet lowers insulin levels and increases insulin sensitivity, allowing excess fluid to be released. These combined effects are the origin of the claim that the weight lost on keto diets is due to water loss. In the very beginning, this is true, but subsequently, of course, it is not." Where is the hustling here?

Athletic performance

Alan suffered from keto-adaptation, and though he appeared to know about it, citing Phinney, he still complained about his athletic performance, which is known to take 4-6 weeks to resolve in professional athletes. He didn't plan to do the diet for that length of time, so it seems disingenuous to bring this up as an issue.

Give it more time and have a good reason

As I've outlined in my post about how to try a carnivorous diet (Eat meat. Not too little. Mostly Fat), It takes about 30 days to see appreciable differences from this diet (and often more), compared to a merely ketogenic one. Those differences are really only possible if you've tried a ketogenic diet and didn't get everything you hoped for, and if those hopes were within certain categories none of which seem to apply to Alan. I don't advocate eating a carnivorous diet for no reason at all, and even the "unparalleled fuck you" Alan describes is only fun for a little while compared to the fun you could be having by eating whatever you please. It speaks to the power of the intervention that any of us are doing this more than 13 days at all.

End-to-end citations

[Mel1942]

Mellanby, Kenneth. “Metabolic Water and Desiccation.” Nature 150, no. 3792 (July 1942): 21. https://doi.org/10.1038/150021a0.

"The utilization by the body of ingested food substances and of tissue reserves yields among other things quantities of metabolic water. As the complete combustion of 100 gm. of fat produces about 110 gm. of metabolic water, whereas 100 gm. of carbohydrate yields only 55 gm. of water, fat reserves and fatty foods are believed to be particularly valuable as a protection against desiccation. This contention would appear to be supported by the fact that many animals which exist in deserts have large reserves of fat."

[Mur2012]

Murphy, Neil, Teresa Norat, Pietro Ferrari, Mazda Jenab, Bas Bueno-de-Mesquita, Guri Skeie, Christina C. Dahm, et al. “Dietary Fibre Intake and Risks of Cancers of the Colon and Rectum in the European Prospective Investigation into Cancer and Nutrition (EPIC).” PLoS ONE 7, no. 6 (June 22, 2012). https://doi.org/10.1371/journal.pone.0039361.

"After a mean follow-up of 11.0 years, 4,517 incident cases of colorectal cancer were documented. Total, cereal, fruit, and vegetable fibre intakes were estimated from dietary questionnaires at baseline. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using Cox proportional hazards models stratified by age, sex, and centre, and adjusted for total energy intake, body mass index, physical activity, smoking, education, menopausal status, hormone replacement therapy, oral contraceptive use, and intakes of alcohol, folate, red and processed meats, and calcium. After multivariable adjustments, total dietary fibre was inversely associated with colorectal cancer (HR per 10 g/day increase in fibre 0.87, 95% CI: 0.79–0.96). Similar linear associations were observed for colon and rectal cancers. The association between total dietary fibre and risk of colorectal cancer risk did not differ by age, sex, or anthropometric, lifestyle, and dietary variables. Fibre from cereals and fibre from fruit and vegetables were similarly associated with colon cancer; but for rectal cancer, the inverse association was only evident for fibre from cereals."

[War2012]

Warren, John B, Simon Day, and Peter Feldschreiber. “Symmetrical Analysis of Risk–Benefit.” British Journal of Clinical Pharmacology 74, no. 5 (November 2012): 757–61. https://doi.org/10.1111/j.1365-2125.2012.04265.x.

[See graphic]

An Open Dialogue with Alan Levinovitz on Miracle Cures

Introduction for outside readers

Twitter can be a wonderful way to learn from others, but as a medium for in-depth, multifaceted discussion it is painfully difficult. Recently I attempted to start a sincere dicussion with Alan Levinovitz, an author and journalist with expertise in religious studies, who has also written extensively about diet and science.

I have not yet read Alan's book, The Gluten Lie, and so I'm gleaning the following from reviews. I trust he will correct my interpretation of his position as needed, but roughly: In the book he argues that eating gluten is rarely a legitimate problem, but the gluten-free movement is instead reflective of a widespread tendency to project notions of good and evil onto food, thus making dietary prescriptions operate more like religious rituals than like scientific therapies. Adding to that the notion that faith is extremely powerful — so powerful that it can emulate "real" treatments — brings him to the conclusion that many people are unnecessarily restricting their diets, because the cure they think they have found is a trick of psychology. Given the potential downsides to food restriction, from nutrient deprivation to social exclusion and missed enjoyment, he aims to dispel myths of this type.

I have great sympathy for many of these ideas. The authors he criticises in his book, for example, William Davis and David Perlmutter, are authors whose reasoning I likewise have found fault with, even though I agree with some of their conclusions despite that reasoning. I also perceive much of the plant-based rhetoric, not just from vegans, but even within the Paleo community, to be largely driven by notions of purity, rather than scientifically based. This has been beautifully articulated by Nick Mailer in his lecture on Paleo Puritanism from AHS16. If you haven't watched it, you're in for a treat. Moreover, though I rarely talk about it here, I've been fascinated with religion, philosophy, psychology, and cognitive neuroscience for decades, much moreso than any of this nutrition stuff. While I am an atheist, I harbour extreme skepticism of certain assumptions often made about the nature of causality and knowledge.

I became aware of Alan's work, because he recently wrote an article criticising the carnivore diet: The Carnivore Diet Is the Latest Fad to Ignore That Food Does More Than Just Feed Us. It is this article that got me talking with Alan on Twitter. After a few rounds of (remarkably respectful) dialogue, he accepted my invitation to an epistolary discussion in particular of the following idea:

(When) does contesting a solution to a chronic disease entail contesting the presence of the disease itself?

So the purpose of this article is to summarise the brief points we already touched on and to invite Alan to discuss these ideas more fully so that we can come to a mutual understanding of where we differ, or a realisation that we don't. I appreciate the opportunity! It is my intention to make the best good faith argument I can, because I sincerely want to understand this phenomenon. That means I'm going to give the most cogent representation of my understanding of the arguments I'm not yet persuaded by, and seek to avoid even subtly diminishing them in order to make them less persuasive.

For reference, here is a link to the beginning of our discussion on Twitter. https://twitter.com/KetoCarnivore/status/1043586827402133504 I admit to a little testiness in that initial contact, but we got over it. In true Twitter form, the discussion immediately subdivided into multiple threads involving different subsets of commenters, and high potential for sidetracking and confusion.

I also have two objections to the article that don't pertain to this focal question, They both have to do with the keto-adaptation phase of ketogeic diets, not carnivory in particular. I have put them in a separate article so as not to disrupt the dialogue here.

(When) does contesting a solution to a chronic disease entail contesting the presence of the disease itself?

Consider a chronic debilitating disease, such as rheumatoid arthritis. RA is problematic because it has the following properties:

  • Incurable according to conventional medicine, i.e. there is no scientifically accepted cure

    This is a problem because it can be taken as an indication that the probability there is any cure at all is low. Our confidence will be modified by our confidence in medical science generally, and our conception of the possible causes of the disease.

    On the other hand, insofar as the disease is considered incurable, it should elevate the importance of any new findings of people curing it.

  • Can go into remission "spontaneously" (but see below)

    This is a problem because it makes it hard to tell when a therapy just happened to coincide with the remission that was going to occur anyway.

  • Involves difficult to measure, subjective symptoms of pain

    Without good measurement, it's difficult to evaluate the extent of an effect. Moreover, pain is known to be more subject to perceptual manipulation than visible symptoms, and is considered to be more responsive to placebo effect.

  • There are various claims of very different "alternative" therapies curing it, including veganism, carnivory, and homeopathy

    This calls into question plausible mechanisms, and in the case of the mechanism-implausible homeopathy, additionally adds evidence to the susceptibility of the disease in question to modification by faith alone.

    As Alan points out, in this case, the possible consistent interpretations of this situation are:

    • At least some people swearing by the proposals are either lying or mistaken
    • RA has multiple etiologies amenable to different solutions depending on which
    • A common factor or effect in plausible proposals is responsible for the remission, making each vastly unnecessarily restrictive

I want to focus on the possibility that people are mistaken, because I think most of the meat of the problem is here. The way to contest the solution but not the disease is to claim that the solution is spurious. There are two main ways one could make such a mistake. The mistake could be a matter of real remission that is only coincidentally related to the therapy, or the mistake could be due to a psychosomatic effect. Let's consider each in isolation first, with the assumption that it is the only one in operation.

Spontaneous remission

Alan pointed out in his article that RA can go into spontaneous remission. However, my literature search indicates that this is constrained with respect to the course of the disease, and the duration, as well as the initiating context. In the paper he cites [Sha2010], the authors describe how a small minority of RA patients go into "drug-free" remissions. By this phrase, they do not mean remission without drugs! Drug-free remission here means that after successful drug treatment, some patients stay in remission after withdrawal of the drug. In almost half the disease returns. Spontaneous remission without a drug is quite rare, happens only in very early stages of the disease, i.e. within a few months of diagnosis, and lasts only for a few months [Val2008]. It is considered imperative to do whatever it takes to get remission by drug treatment early to prevent ongoing damage, and exactly because later stage disease doesn't typically have such periods of relief.

This gives us a way to quantify the probability that the alternative treatment in question was responsible for the remission, by looking at the mean and variance of the time since diagnosis, and duration of remission. I do not have actual statistics on this for RA, but consider this as a thought experiment. If spontaneous remission rates start, say, at 10% in the first month, fall to 1% after 4 years, and after 8 it's unheard of, then someone going into remission at 4 years after diagnosis is already extraordinarily unlikely, and at 15 it's astronomically so. Similarly, if that kind of remission lasts on average 6 months, with a standard deviation of 4, then a remission of 2 years due to any intervention should be considered highly likely to have something to do with the intervention. No drug study would ever be doubted at that level. Granted, a drug study will have multiple people, and will track everyone who tries it, not just those who succeed. Nonetheless, to ignore a statistical anomaly of that magnitude smacks of strong, and possibly irrational bias against the intervention.

I further offer the idea that multiple trials off and on the intervention can narrow down the uncertainty. If the patient can replicate the effect over and over by stopping and starting the intervention, then it uncouples the potential for coincidence. All of this assumes, of course, that the mechanism of action of the therapy is not placebo or faith. So let's move on to that possibility.

What are the limits of placebo and faith healing?

Placebo

Where I think Alan and I might differ the most is in our belief about how much effect can possibly be attributed to belief. (Don't think about that last sentence too hard!) One type of effect that is considered to reflect this kind of mind power is the placebo in a controlled trial. However, placebo effects are so limited that several people have questioned their extent or even existence, noting that the level of variance that can be achieved in practice and that is normally attributed to placebo, can be mostly or completely explained by regression to the mean [McD1983], [Hró2001], [Sen2009], [Col2015].

Regression to the mean describes the tendency of measurements that have some variability, when taken multiple times, to stay close to the average. In particular, if you get a measurement that is extreme (far from average), there is a high probability that a second measurement later will be less extreme. For example, suppose that I have a particularly bad flare up of RA symptoms this month. It's highly likely that next month my symptoms will be less severe even if I do nothing. This scenario is of clinical relevance because patients in trials are likely to have been enrolled because they are suffering more than others. See [Yud1996] for a discussion of the problem and how to account for it.

In meta-analysis, placebo effects have mostly not been found clinically significant [Hró2011], . An important potential exception that is relevant here is pain, which along with nausea has varied enough to reach significance, though arguably accountable through biased reporting [Hró2011]. I'd also conjecture that pain, like other sensations, is not perceived by people in a linear fashion (see e.g. Weber's and Fechner's laws for perception of light and sound intensity). This may make it difficult to scale based on people's reports when asked to rate pain.

What the statisticians who deny that the placebo effect is psychosomatic are saying is that you don't need a psychological explanation to account for the fact that people in the control group also get better! The statistical properties of the situation already anticipate this effect. Critically, though, this kind of statistical effect can only happen within mathematically definable limitations. If the average measurement is 100, and the standard deviation is 5, then regression to the mean wouldn't account for a drop to 50.

Stress

Besides the placebo effect, the idea came up in our Twitter discussion that "stress" can cause psychosomatic symptoms. My issue with this idea is simply this: emotions are physiological. Feeling stress can be a symptom of physiological disturbance. It may appear quickly, relative to other bodily symptoms, but just because it comes first doesn't mean that it caused the bodily symptoms. I see no way to systematically distinguish between a case where a food caused, for example (and I don't mean to suggest this pathway is definitive), an immune response which results in inflammation of the brain which influences affect in the form of increased feelings of stress, followed by a rash, pain, or gastrointestinal distress. To claim that the bodily symptom was caused by stress is making an unwarranted assumption.

In my own personal experience of psychiatric disease (I have latent type 2 bipolar disorder that went into remission following a dietary change), I have suffered life situations far more stressful since my dietary changes than during the worst years of the disease: divorce, psychological abuse in a subsequent long-term relationship, two moves, a new job followed by leaving the job to work for myself (two career changes), oldest child left home, drastic changes in financial position, and taking on a mortgage for the first time. These are considered to be among the most stressful events a person can experience. Yet, my mental and physical health have continued to be superior to the prior 5 years when my bipolar disorder was at its peak and my life was stable. I think the diet has made me more resilient to stress and more physically healthy. Someone else might argue that it made me more resilient to stress and consequently more healthy. What it didn't do was simply reflect the environmental stress in my life.

Faith

Is it possible for a person to cure a disease simply by believing in it strongly enough? My first experience with this idea was in my first year at university. I was studying mathematics, and there was an extremely mathematically intelligent, stereotypically nerdy young woman in my classes. We sometimes studied together, and later, while my life was falling apart due to my first depressive episodes, she was going on to become a talented cryptographer. One of the things she told me was that she found Jesus as a child and that He cured her epilepsy. I did and do find this astounding, and I've never forgotten it. At the time, I absolutely believed that her experience was real, and that it was a testament to the power of the mind.

What makes this not scientifically palatable? Moreover, why isn't everyone falling over themselves to get in His good graces? I think the only reason this kind of effect is dismissed is because of numbers. I'm not really familiar with the literature, but as far as I understand, this kind of result simply hasn't been reproducible at scale. Religious believers, as far as I'm aware, are not better off generally than non-believers in their rate of suffering disease. The only studies I've seen in which prayer made any difference were, again, interestingly, in pain, but they were weak. For example, Matthews found that when patients with RA (no less) were prayed for, there was a (barely) statistically significant effect on 2 of 10 measures of symptoms after 6 months [Mat2000]. This effect was not replicated when patients were prayed for at a distance without their knowledge. Similarly, there was a study showing some relief of migraine pain from prayer [Taj2017]. The effect was significant, but modest, in that the participants were still in a lot of pain at the end. This was not a cure, by any means. Moreover, the baseline differences between groups increased the likelihood in regression to mean effects in the treatment group, compared to the control.

When people pray for their own health, the main effect seems to be on one's ability to cope with the problems, not the issues themselves. Hollywell et al. did a meta analysis finding that there was a correlation between prayer and better mood, but also a correlation between chronic disease and prayer [Hol2008]. While there is an obvious explanation for the increased prayer in that situation, it also suggests it isn't helping much. In a similar study, Anderson also found that prayer appears to help with coping with chronic disease, but had no effect on outcomes [And2016].

This lack of replicability doesn't mean that faith healing doesn't happen in individuals, but it does raise questions about what is so difficult to replicate that only a select few can manage to do it. I would think that if you believed in this kind of power you would be searching for any way to replicate it possible.

The point of a controlled trial, and particularly blinding, is to distinguish between the effects that come from a treatment and the effects that come from expectation of being healed. If we are going to chalk up the effects of the carnivore diet to a strong expectation, then we have to explain at least the following things.

First, back when the population of people trying this diet was much smaller than it is now, no one had heard of it curing psychiatric or autoimmune disorders. People came to "Zero Carb" for weight loss. Then people started sharing stories about other health issues surprisingly resolving. When my mood disorder appeared to vanish, no one was more surprised than I was. This effect could not have happened due to my faith in the ability of the diet to have that effect. It wasn't on my radar of possibilities. I didn't even necessarily think it was a particularly healthy diet! I thought of it as a temporary weight loss manoeuver. Here I think the only way that we can argue that the current collection of autoimmune and psychiatric effects of a carnivore diet are due to faith is to do some kind of bootstrapping. We must argue that these initial first examples were mistakes of the chance type that started to influence newcomers, gathering their own momentum, and shaping subsequent expectations in a feedback loop.

Second, the people who now come to try a carnivore diet are for the most part people for whom traditional medical solutions didn't work. How do we explain the fact that these people couldn't muster the required faith in a traditional medicine, but could muster it with diet? To use my own story as an example again, I was practically overjoyed when my depressive disorder was rediagnosed as bipolar disorder, because I was sure that now, finally, I would have access to the right medicine to help me. It didn't help me; not even a little bit. My bipolar disorder was progressing.

So to return to the orginal subject: I think Alan is saying that he can question a solution to a disease without questioning the presence of the disease, by saying that either the disease wasn't cured by the treatment, it just happened to remit at the same time, or it was cured by faith, which has nothing to do with the purported biological mechanisms of the therapy. Moreover, the specifics of the therapy are arbitrary, and need only meet certain requirements that make us able enact the faith healing mechanism. I would like to learn from Alan what he thinks these requirements are.

Plausible mechanisms and prior probabilities

It's not enough for the supposed mechanisms to be wrong for a therapy to be discounted. Observations hold whether we understand them or not. This is the scientific process: recognise patterns in observations and figure out what accounts for them. If you have to overturn a prior held belief to explain the facts then so be it. If you later find out that your belief in why a therapy works isn't scientifically tenable, it doesn't negate the observations.

However, I agree with Alan that if there are no plausible mechanisms for a therapy, then it is implausible. For example, I cannot fathom any mechanism by which homeopathy would work. If it were the case that many people were testifying that homeopathy reduced their RA symptoms to nothing, rather than a regression to the mean result of small reductions in pain for short periods, then I would have to concede that either there is something that homeopathy is doing that we physically don't yet understand, or that homeopathy dials in to the requirements of faith healing in a way that even prayer cannot. However, I don't think we're in that situation as far as homeopathy goes.

On the other hand, there are plausible mechanisms for a carnivore diet to heal RA, related to intestinal permeability. Some plant compounds are known to increase intestinal permeability (See e.g. [Mil2011], [Shi2013]). At the same time an important function of the intestinal barrier is to prevent the absorption of toxic plant secondary metabolites [Fin2010]. Autoimmune conditions are associated with intestinal barrier dysfunction [Smi1985], [Jul2016] which some argue is causal [Arr2006], though others attribute this to the treatment of NSAIDs (e.g. [Bja1984]). Celiac disease, also connected to intestinal permeability, is associated with RA [Vol2006], [Ler2015], [War2015]. So intestinal permeability may on the one hand be worsened by the intake of certain plants, and simultaneously such intake may worsen the autoimmune condition itself because of toxic exposure in a way that it couldn't in someone with an intact barrier. Not all plant compounds are implicated equally, and the criticism that carnivory may often overshoot what is necessary in an individual case is a valid one. Nonetheless, I think this knowledge should give us a higher prior probability when considering the possibility that this is a "real" solution than we would give homeopathy.

My perception is that the vehement criticism against the carnivore diet, not necessarily by Alan, but in general, is that most people have a strong prior belief that plant foods are not only harmless, but in fact beneficial. This position does not stand up to scientific scrutiny. There has never been benefit demonstrated in a study that was not either epidemiological, or based on extrapolation from drug-like effects in concentrated isolates. In neither case was there a comparison between humans eating only animal based foods, and those eating animal based foods with the addition of plants.

Plants are toxic to their predators as a biological strategy. It is pure fantasy to imagine that we developed in happy symbiosis. To the contrary, it is an arms race. Our resistance to such toxins is well honed, but limited, and it is reasonable to expect that even the plants of lowest toxicy which we selectively bred for our own use could result in harm if our resistance were lowered.

What I have been seeing for nearly a decade since I started on this diet is a dismissal of carnivory based on it sounding implausible to people because of their prior convictions about plants. That is, the objections don't come because they've looked at the difference between a diet with and without plants, and conclude that there is no discernible metabolic or biological effect of the difference, in the way that there is no discernible effect between a homeopathic remedy and a vial of water. To the contrary, all kinds of things must be happening differently. Rather, most people assume that there are differences, but that these differences have to be detrimental, and couldn't possibly be therapeutic. This is a double standard.

What's the difference?

It seems to me that when you characterise what someone offers as the explanation for their recovery from a chronic disease as a mistake, you are either claiming that they are a careless or unscientific thinker (because they do not understand the statistics around recovery from their disease), or that they are fooling themselves in a profound way. I concede that this is not the same as questioning the existence of the disease, but it does entail a claim that the disease was in their own power all along. I find it hard to distinguish the claim that I have the power to stop my chronic disease by evoking the right state of mind, and that I am creating my disease by not thereby stopping it. In other words, the disease is a product of my lack of belief. It's "in my head".

Sincerely,

I hope I have generated something of interest for Alan to respond to. I could never have done this on Twitter! I am looking forward to hearing from him.

Amber

End-to-end citations

[And2016]

Anderson, James W., and Paige A. Nunnelley. “Private Prayer Associations with Depression, Anxiety and Other Health Conditions: An Analytical Review of Clinical Studies.” Postgraduate Medicine 128, no. 7 (October 2, 2016): 635–41. https://doi.org/10.1080/00325481.2016.1209962.

"This analytical review evaluated research studies examining the association between private prayer and health conditions. Private prayer was defined as praying for one’s own health. Forty-one articles reported relationships between prayer and health. Studies were assigned prayer frequency scores from 1 to 8 depending on the specificity and frequency of private prayer. Prayer frequency scores ranged from 8 (high specificity of prayer and frequency) to 1 (low specificity of prayer or unstated frequency). Twenty-one studies with prayer scores from 5 to 8 were evaluated in detail. Frequent private prayer was associated with a significantly lower prevalence of depression (P <0.01);higherlevelsof optimism or coping (P < 0.01); and higher levels of mental health (P < 0.01). No significant associations were reported between private prayer and physical heal th or blood pressure. Prospective controlled clinical trials are required to more critically asses the associations between private prayer and health conditions."

[Arr2006]

Arrieta, M C. “Alterations in Intestinal Permeability.” Gut 55, no. 10 (October 1, 2006): 1512–20. https://doi.org/10.1136/gut.2005.085373.

"For decades a variety of pathological states have been associated with abnormal permeability. Many of these are a consequence of intestinal epithelial damage that is associated with disease but not involved in a causal manner in the genesis of disease. However, in several autoimmune conditions it appears that increased permeability is a constant and early feature of the disease process. Furthermore, it is becoming increasingly apparent that in some conditions increased permeability is critical to the development of disease as if it is abrogated the disease does not develop. This is particularly true in type 1 diabetes. In other diseases such as Crohn’s disease or coeliac disease, a similar pattern of findings are apparent but the experiment to try and prevent disease by preventing the increase in permeability has not been performed."

[Bja1984]

Bjarnason, Ingvar, Alex So, A.Jonathan Levi, TimothyJ. Peters, Peter Williams, GiuseppeD. Zanelli, J.Michael Gumpel, and Barbara Ansell. “INTESTINAL PERMEABILITY AND INFLAMMATION IN RHEUMATOID ARTHRITIS: EFFECTS OF NON-STEROIDAL ANTI-INFLAMMATORY DRUGS.” The Lancet 324, no. 8413 (November 1984): 1171–74. https://doi.org/10.1016/S0140-6736(84)92739-9.

"The suggestion that the intestinal mucosa may be abnormally permeable and a site of absorption of antigens in rheumatoid arthritis was tested by the use of a 51Cr-EDTA (edetic acid) absorption test. 24 patients with rheumatoid arthritis excreted significantly more 51Cr-EDTA than did 34 controls. Intestinal permeability was normal in untreated patients but almost invariably abnormal in patients treated with non-steroidal anti-inflammatory drugs. Studies in patients with osteoarthritis showed that the permeability abnormalities were due to an effect of NSAIDs on both the proximal and the distal intestine and that the effect was systemically mediated. Indium-111-labelled leucocyte scans showed ileocaecal inflammation in 6 of 9 patients on or recently on NSAIDs. Although increased intestinal permeability does not seem to be important in the pathogenesis of rheumatoid arthritis, the administration of NSAIDs may lead to loss of intestinal integrity, thus facilitating antigenic absorption and perhaps contributing to persistence of the disease."

[Col2015]

Colquhoun, David. “Placebo Effects Are Weak: Regression to the Mean Is the Main Reason Ineffective Treatments Appear to Work.” DC’s Improbable Science, December 11, 2015. http://www.dcscience.net/2015/12/11/placebo-effects-are-weak-regression-to-the-mean-is-the-main-reason-ineffective-treatments-appear-to-work/.

"[T]he placebo effect, though a real phenomenon, seems to be quite small. In most cases it is so small that it would be barely perceptible to most patients. Most of the reason why so many people think that medicines work when they don’t isn’t a result of the placebo response, but it’s the result of a statistical artefact.

"Regression to the mean is a potent source of deception

"The get-better-anyway effect has a technical name, regression to the mean. It has been understood since Francis Galton described it in 1886 (see Senn, 2011 for the history). It is a statistical phenomenon, and it can be treated mathematically (see references, below). But when you think about it, it’s simply common sense."

[Fas2012]

Fasano, Alessio. “Leaky Gut and Autoimmune Diseases.” Clinical Reviews in Allergy & Immunology 42, no. 1 (February 2012): 71–78. https://doi.org/10.1007/s12016-011-8291-x.

"Autoimmune diseases are characterized by tissue damage and loss of function due to an immune response that is directed against specific organs. This review is focused on the role of impaired intestinal barrier function on autoimmune pathogenesis. Together with the gut-associated lymphoid tissue and the neuroendocrine network, the intestinal epithelial barrier, with its intercellular tight junctions, controls the equilibrium between tolerance and immunity to non-self antigens. Zonulin is the only physiologic modulator of intercellular tight junctions described so far that is involved in trafficking of macromolecules and, therefore, in tolerance/ immune response balance. When the zonulin pathway is deregulated in genetically susceptible individuals, autoim- mune disorders can occur. This new paradigm subverts traditional theories underlying the development of these diseases and suggests that these processes can be arrested if the interplay between genes and environmental triggers is prevented by re-establishing the zonulin-dependent intestinal barrier function. Both animal models and recent clinical evidence support this new paradigm and provide the rationale for innovative approaches to prevent and treat autoimmune diseases."

[Fin2010]

Fink-Gremmels, Johanna. “Defense Mechanisms against Toxic Phytochemicals in the Diet of Domestic Animals.” Molecular Nutrition & Food Research 54, no. 2 (February 2010): 249–58. https://doi.org/10.1002/mnfr.200900361.

"or many decades, the epithelial cell layer, closely connected by tight junction proteins, has been described as the barrier system of the intestines, preventing the absorption of many [Plant Secondary Metabolites], as well as biocides and drugs. This assumption was supported by the identification of specific transporters that facilitate the absorption of polar nutrients, such as amino acids and sugars (soluble ligand carriers previously denoted as organic anion transporters, glucose transporters, and others) [50]."

[Hol2008]

Hollywell, Claire, and Jan Walker. “Private Prayer as a Suitable Intervention for Hospitalised Patients: A Critical Review of the Literature.” Journal of Clinical Nursing, November 2008. https://doi.org/10.1111/j.1365-2702.2008.02510.x.

"Summary of key findings

  • Prayer, measured by frequency, is usually associated with lower levels of depression and anxiety. (a) But most of the studies that show positive associations between prayer and wellbeing were located in areas that have strong Christian traditions and involved samples that report relatively high levels of religiosity, church attendance and use of prayer.
  • Church attenders, older people, women, those who are poor, less well educated and have chronic health problems make more frequent use of prayer. (a) It may be that the weak and vulnerable in society are more likely to turn to the church and to prayer in times of difficulty.
  • Prayer is a coping action that mediates between religious faith and wellbeing.
  • Prayer takes different forms, some beneficial, others possibly not: (a) Devotional prayers that take the form of an intimate dialogue with a supportive God are associated with improved optimism, wellbeing and function. (b) Prayers that involve only pleas for help in extremis may, in the absence of a pre-existing faith, be associ- ated with increased distress and possibly poorer function"
[Hró2001]

Hróbjartsson, Asbjørn, and Peter C. Gøtzsche. “Is the Placebo Powerless?” New England Journal of Medicine 344, no. 21 (May 24, 2001): 1594–1602. https://doi.org/10.1056/NEJM200105243442106.

"We found little evidence in general that placebos had powerful clinical effects. Although placebos had no significant effects on objective or binary outcomes, they had possible small benefits in studies with continuous subjective outcomes and for the treatment of pain. Outside the setting of clinical trials, there is no justification for the use of placebos."

[...]

"Placebos have been reported to improve subjective and objective outcomes in up to 30 to 40 percent of patients with a wide range of clinical conditions, such as pain, asthma, high blood pressure, and even myocardial infarction.1-3 In his 1955 article “The Powerful Placebo,” Beecher concluded, “It is evident that placebos have a high degree of therapeutic effectiveness in treating subjective responses, decided improvement, interpreted under the unknowns technique as a real therapeutic effect, being produced in 35.2±2.2% of cases.[1]”

"Beecher's article and the 35 percent figure are often cited as evidence that a placebo can be an important medical treatment. The vast majority of reports on placebos, including Beecher's article, have estimated the effect of placebo as the difference from base line in the condition of patients in the placebo group of a randomized trial after treatment. With this approach, the effect of placebo cannot be distinguished from the natural course of the disease, regression to the mean, and the effects of other factors.4-6 The reported large effects of placebo could therefore, at least in part, be artifacts of inadequate research methods."

[Hró2011](1, 2)

Hróbjartsson, Asbjørn, and Peter C. Gøtzsche. “Placebo Interventions for All Clinical Conditions.” The Cochrane Database of Systematic Reviews, no. 1 (January 20, 2010): CD003974. https://doi.org/10.1002/14651858.CD003974.pub3.

"We did not find that placebo interventions have important clinical effects in general. However, in certain settings placebo interventions can influence patient-reported outcomes, especially pain and nausea, though it is difficult to distinguish patient-reported effects of placebo from biased reporting. The effect on pain varied, even among trials with low risk of bias, from negligible to clinically important. Variations in the effect of placebo were partly explained by variations in how trials were conducted and how patients were informed."

[Jul2016]

Julio-Pieper, M., and J.A. Bravo. “Intestinal Barrier and Behavior.” In International Review of Neurobiology, 131:127–41. Elsevier, 2016. https://doi.org/10.1016/bs.irn.2016.08.006.

"The intestinal barrier function contributes to gut homeostasis by modulating absorption of water, electrolytes, and nutrients from the lumen into the circulation while restricting the passage of noxious luminal substances and microorganisms. Chronic conditions such as rheumatoid arthritis, inflammatory bowel disease, and celiac disease are associated to intestinal barrier dysfunction. Here, the hypothesis is that a leaky intestinal wall allowing for indiscriminate passage of intraluminal compounds to the vascular compartment could in turn lead to systemic inflammation. An increasing number of studies are now investigating the association between gut permeability and CNS disorders, under the premise that translocation of intestinal luminal contents could affect CNS function, either directly or indirectly. Still, it is unknown whether disruption of intestinal barrier is a causative agent or a consequence in these situations. Here, we discuss the latest evidence pointing to an association between increased gut permeability and disrupted behavioral responses."

[Ler2015]

Lerner, Aaron, and Torsten Matthias. “Rheumatoid Arthritis–Celiac Disease Relationship: Joints Get That Gut Feeling.” Autoimmunity Reviews 14, no. 11 (November 2015): 1038–47. https://doi.org/10.1016/j.autrev.2015.07.007.

"Rheumatoid arthritis (RA) and celiac disease (CD) belong to the autoimmune disease family. Despite being separate entities they share multiple aspects. Epidemiologically they share comparable incidence environmental influences, associated antibodies and a recent incidental surge. They differ in their HLA pre-dispositions and specific predictive and diagnostic biomarkers. At the clinical level, celiac disease exhibits extra-intestinal rheumatic manifestations and RA gastrointestinal ones. Small bowel pathology exists in rheumatic patients. A trend towards responsiveness to a gluten free diet has been observed, ameliorating celiac rheumatic manifestations, whereas dietary interventions for rheumatoid arthritis remain controversial. Pathophysiologically, both diseases are mediated by endogenous enzymes in the target organs. The infectious, dysbiotic and increased intestinal permeability theories, as drivers of the autoimmune cascade, apply to both diseases. Contrary to their specific HLA pre-disposition, the diseases share multiple non-HLA loci. Those genes are crucial for activation and regulation of adaptive and innate immunity. Recently, light was shed on the interaction between host genetics and microbiota composition in relation to CD and RA susceptibility, connecting bugs and us and autoimmunity. A better understanding of the above mentioned similarities in the gut-joint inter-relationship, may elucidate additional facets in the mosaic of autoimmunity, relating CD to RA."

[Mat2000]

Matthews, DA. “Effects of Intercessory Prayer on Patients with Rheumatoid Arthritis. - PubMed - NCBI.” Accessed October 2, 2018. https://www.ncbi.nlm.nih.gov/pubmed/11142453.

"Multivariate analysis showed significant overall improvement in the 10 outcome variables over a 6-month postintervention follow-up period for group 1 (n=26) when compared with a 6-month preintervention follow-up period for the waiting-list control group (group2 n=14) (P > .0001). Univariate analysis (Table 3) showed a greate reduction over the 6-month follow-up period for group 1 versus group 2 in the mean number of temder joints (differences of -7.1 vs 0.5, P = .016) and for MHAQ scores (differnces of -3.6 vs 1.5, P= 0.4)"

[McD1983]

McDonald, Clement J., Steven A. Mazzuca, and George P. McCabe. “How Much of the Placebo ‘Effect’ Is Really Statistical Regression?” Statistics in Medicine 2, no. 4 (October 1983): 417–27. https://doi.org/10.1002/sim.4780020401.

"Statistical regression to the mean predicts that patients selected for abnormalcy will, on the average, tend to improve. We argue that most improvements attributed to the placebo effect are actually instances of statistical regression. First, whereas older clinical trials susceptible to regression resulted in a marked improvement in placebo-treated patients, in a modern'series of clinical trials whose design tended to protect against regression, we found no significant improvement (median change 0.3 percent, p > 0.05) in placebo-treated patients."

[Mil2011]

Milner, Sinead Eileen, Nigel Patrick Brunton, Peter Wyn Jones, Nora Mary O’ Brien, Stuart Gerard Collins, and Anita Rose Maguire. “Bioactivities of Glycoalkaloids and Their Aglycones from Solanum Species.” Journal of Agricultural and Food Chemistry 59, no. 8 (April 27, 2011): 3454–84. https://doi.org/10.1021/jf200439q.

"Glycoalkaloids, a class of nitrogen-containing steroidal glycosides, are biologically active secondary plant metabolites and are commonly found in plants of the Solanum genus. These include many common vital agricultural plants including potato ( Solanum tuberosum ), tomato ( Solanum lycopersicum ), and aubergine ( Solanum melongena )."

"The toxicity of glyoalkaloids in humans is well documented, with “ solanine ” poisoning from blighted, green, or sprouted potatoes being reported as early as 1980. 14 - 16,27 - 29 Accordingly, current safety regulations limit their content in the edible tuber to 20 mg/100 g fresh weight (fw). 30,31 The mechanism of toxicity induced by glycoalkaloids is associated with their membrane-disruptive properties 32 - 34 and their inhibition of acetylcholinesterase activity. 35 - 39 Bioactivity of glycoalkaloids is not limited to their toxicity; they have been reported to possess anticancer, anticholesterol, and anti-inflammatory properties, for example, and some of these e ff ects have been reviewed. 1,14,16,30,40 - 47 Although there have been multiple reviews of individual glycoalkaloids in the literature, including separate reviews of potato 45 or tomato 1 glycoalkaloids, this is the fi rst review to examine and compare the bioactivities, toxicities, and synergisms of action of the principal Solanum glycoalkaloids. Furthermore, the mechanisms of action are discussed, and the relationship between molecular structure and bioactivity profile is presented."

[Sen2009]

Senn, Stephen. “Three Things That Every Medical Writer Should Know about Statistics” 18, no. 3 (2009): 5.

"Regression to the mean is the tendency for members of a population who have been selected because they are extreme to be less extreme when measured again [4, 5]. Because entry into clinical trials is usually only allowed if patients have extreme values (diastolic blood pressure above 95 mmHg, Hamilton depression score greater than or equal to 22, forced expiratory volume in one second less than 75% of predicted etc.), regression to the mean is a phenomenon that is likely to affect many clinical trials. We can expect that patients will appear to improve even if the treatment is ineffective. Regression to the mean is a plausible explanation, for example, for the ‘placebo effect’ which then becomes, as I hope to explain, a purely statistical rather than psychological phenomenon."

[Sha2010]

Shammas, Rania M., Veena K. Ranganath, and Harold E. Paulus. “Remission in Rheumatoid Arthritis.” Current Rheumatology Reports 12, no. 5 (October 2010): 355–62. https://doi.org/10.1007/s11926-010-0121-2.

"Spontaneous remission is not uncommon in patients who present with very early arthritis, some of whom may meet criteria for RA over less than a few months. Spontaneous remission is thought of as a “natural remission,” in which disease activity essentially disappears, and medications are no longer required. Spontaneous remission may be seen in 13% to 55% of individuals presenting with undifferentiated arthritis, probably as a result of different underlying etiologies, such as a transient viral infection [32]. About one third of patients with undifferentiated arthritis go on to develop RA."

[Shi2013]

Shiobara, Tomoko, Takeo Usui, Junkyu Han, Hiroko Isoda, and Yoko Nagumo. “The Reversible Increase in Tight Junction Permeability Induced by Capsaicin Is Mediated via Cofilin-Actin Cytoskeletal Dynamics and Decreased Level of Occludin.” PLOS ONE 8, no. 11 (November 18, 2013): e79954. https://doi.org/10.1371/journal.pone.0079954.

"Previous results demonstrated that capsaicin induces the reversible tight junctions (TJ) opening via cofilin activation. The present study investigated the mechanisms underlying the reversible TJ opening and compared the effect to the irreversible opening induced by actin inhibitors. Capsaicin treatment induced the F-actin alteration unique to capsaicin compared to actin-interacting agents such as latrunculin A, which opens TJ irreversibly. Along with TJ opening, capsaicin decreased the level of F-actin at bicellular junctions but increased it at tricellular junctions accompanied with its concentration on the apical side of the lateral membrane. No change in TJ protein localization was observed upon exposure to capsaicin, but the amount of occludin was decreased significantly. In addition, cosedimentation analyses suggested a decrease in the interactions forming TJ, thereby weakening TJ tightness. Introduction of cofilin, LIMK and occludin into the cell monolayers confirmed their contribution to the transepithelial electrical resistance decrease. Finally, exposure of monolayers to capsaicin augmented the paracellular passage of both charged and uncharged compounds, as well as of insulin, indicating that capsaicin can be employed to modulate epithelial permeability. Our results demonstrate that capsaicin induces TJ opening through a unique mechanism, and suggest that it is a new type of paracellular permeability enhancer."

[Smi1985]

Smith, M. D., R. A. Gibson, and P. M. Brooks. “Abnormal Bowel Permeability in Ankylosing Spondylitis and Rheumatoid Arthritis.” The Journal of Rheumatology 12, no. 2 (April 1985): 299–305.

"Intestinal permeability was measured using a low molecular weight polyethylene glycol as a permeability marker in patients with osteoarthritis, ankylosing spondylitis (AS) and rheumatoid arthritis (RA). Patients with AS showed a significant increase in bowel permeability when compared to controls. Intestinal permeability was also increased in patients with active RA but was less than the control group in RA patients who did not have active joint disease."

[Taj2017]

Tajadini, Haleh, Nasser Zangiabadi, Kouros Divsalar, Hossein Safizadeh, Zahra Esmaili, and Hossein Rafiei. “Effect of Prayer on Intensity of Migraine Headache.” Journal of Evidence-Based Complementary & Alternative Medicine 22, no. 1 (January 2017): 37–40. https://doi.org/10.1177/2156587215627551.

"At the beginning of study and before intervention, the mean score of pain in patients in groups A and B were 5.7 ± 1.6 and 6.5 ± 1.9, respectively. According to results of independent t test, mean score of pain intensity at the beginning of study were similar between patients in 2 groups (P > .05). Three month after intervention, mean score of pain intensity decreased in patients in both groups. At this time, the mean scores of pain intensity were 5.4 ± 1.1 and 4.2 ± 2.3 in patients in groups A and B, respectively. This difference between groups was statistically significant (P < .001)."

[Val2008]

Valesini, Guido, Manuela Di Franco, Francesca Romana Spinelli, and Rossana Scrivo. “Induction of Remission in Rheumatoid Arthritis: Criteria and Opportunities.” Rheumatology International 29, no. 2 (December 2008): 131–39. https://doi.org/10.1007/s00296-008-0699-0.

"[R]emission represents the primary objective of treatment in rheumatoid arthritis (RA). In the last two decades, rheumatologists have realized that complete (or almost complete) control of th e disease is essential to prevent long-term joint damage and remission in RA is now more feasible than in the past, thanks to the introduction of biologic disease modifying anti-rheumatic drugs (DMARDs) in RA therapy."

"[P]eople with RA may undergo spontaneous remissions in the early phases of the disease, generally only for a few months."

[Vol2006]

Volta, U, G Deangelis, A Granito, N Petrolini, E Fiorini, M Guidi, P Muratori, and F Bianchi. “Autoimmune Enteropathy and Rheumatoid Arthritis: A New Association in the Field of Autoimmunity.” Digestive and Liver Disease 38, no. 12 (December 2006): 926–29. https://doi.org/10.1016/j.dld.2006.02.003.

"We report the case of a 35-year-old woman with a diagnosis of coeliac disease at the age of 32 due to a severe malabsorption and flat mucosa without endomysial and tissue transglutaminase antibodies. The lack of clinical and histological improvement after 1 year of a gluten-free diet led to a diagnosis of refractory sprue. She had a good clinical response to steroids that were stopped after 3 months when she became pregnant. After delivery, she again started to complain of malabsorption with arthritis. Positivity for enterocyte autoantibodies together with a flat mucosa persistence allowed to identify a condition of autoimmune enteropathy; moreover, a rheumatological assessment gave evidence of an associated rheumatoid arthritis. Treatment by steroids and methotrexate brought to the remission of intestinal and articular symptoms together with an improvement of duodenal histology. This is the first description of an autoimmune enteropathy associated with rheumatoid arthritis. Autoimmune enteropathy should be always ruled out in patients with a villous atrophy unresponsive to a gluten-free diet, autoimmune manifestations and negativity of coeliac disease markers."

[War2015]

Warjri, Synrang Batngen, Tony Ete, Taso Beyong, Bhupen Barman, Kyrshanlang G. Lynrah, Hage Nobin, and Obang Perme. “Coeliac Disease With Rheumatoid Arthritis: An Unusual Association.” Gastroenterology Research 8, no. 1 (February 2015): 167–68. https://doi.org/10.14740/gr641w.

"Coeliac disease has a significant association with many autoimmune disorders. It shares many common genetic and immunological features with other autoimmune diseases. Gluten, a gut-derived antigen, is the driver of the autoimmunity seen in coeliac disease. The altered intestinal permeability found in coeliac patients, coupled with a genetic predisposition and altered immunological response, may result in a systemic immune response that is directed against sites other than the gut. Gut-derived antigens may have a role in the pathogenesis of other autoimmune disorders including rheumatoid arthritis. Here we report a case of adult coeliac disease associated with rheumatoid arthritis."

[Yud1996]

Yudkin, P.L., and I.M. Stratton. “How to Deal with Regression to the Mean in Intervention Studies.” The Lancet 347, no. 8996 (January 1996): 241–43. https://doi.org/10.1016/S0140-6736(96)90410-9.

"RTM is, in essence, a chance finding masquerading as a real one. Despite the name, it is not confined to regression analysis but turns up in several different situations[2]. This paper is concerned with intervention studies. Such studies often target individuals who have unusually high values of a risk variable, such as cholesterol. In a group of such individuals, the mean cholesterol level will, on remeasurement, be lower than the starting mean, even without any intervention or treatment[3-6]. This is the RTM effect. In a group selected on the basis of unusually low values of a variable RTM works in the other direction. RTM occurs with any variable that fluctuates within an individual, either genuinely or due to measurement error[7]. Examples are cholesterol, blood pressure, and plasma glucose. A single measurement may be higher or lower than the individual’s long-term average (or "true") value, but individuals selected because they have a single high measurement will include a disproportionate number whose measurement was by chance higher than their true value. This can be better understood from the figure. The population distribution of the measurement in question consists of numerous individual distributions, with the peak of each representing the individual’s true value. Above any high cut-off point (the vertical line at L) there will be more individuals like A (with their long-term average below the line) than like B (with their long-term average above it). On remeasurement, the values recorded for these individuals will tend to revert towards their long- term average, so that the mean measurement of the group as a whole will fall. In this paper we illustrate the magnitude of the RTM effect in a group of individuals with high cholesterol concentrations, selected from a population with repeated cholesterol measurements and no intervention. We then describe two methods for estimating the RTM effect in an intervention study, and go on to show how intervention studies can be designed specifically to limit the RTM problem."