Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Comment author: Lumifer 03 June 2015 03:29:24PM 5 points [-]

I would recommend a SSC post. You might also want to look at CrazyMeds.

Comment author: michael_b 03 June 2015 08:31:18PM *  1 point [-]

Upvote for interesting and relevant links, although this part made me want to shout at my screen.

(9). Therefore, we should give up on medication and use psychotherapy instead Makes sense right up until you run placebo-controlled trials of psychotherapy ... Another study by the same team finds psychotherapy has an effect size of 0.22 compared to antidepressants’ 0.3-0.5

Even if this is true I don't agree with the cost-benefit analysis. Psychotherapy costs time and money but probably won't cause weight gain, sexual dysfunction and crippling withdrawal if you miss a dose or need to cycle off of them.

EDIT: I guess he says as much in a different article. Hmph.

Systemic review of antidepressants vs placebo commentary

0 michael_b 03 June 2015 12:40PM

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4172306/pdf/zfp_222_3_128.pdf

The background on this story is a community of science people found a bunch of unpublished studies that, when weighed with the studies which supported antidepressant approval, showed they were no more effective than placebo in mild-moderate cases.

Except unlike placebo, antidepressants express a wider range of severe side effects, like worsening depression and suicide.

Isn't this a scandal?  How do psychiatrists still prescribe these en masse?

Comment author: zedzed 10 April 2015 12:50:49PM 0 points [-]

This discussion has already happened at great length here.

To summarize my stance: there's risks, but considering that everyone I've read on discourse.soylent.me has had positive results across the board, from body composition to semen taste. I get noticeably improved mental clarity (along with getting so lean I'd be scared I was undereating if I didn't know precisely how many calories I was eating and clearer skin), which makes me willing to accept those risks. Also, because soylent might be safe and come with a load of benefits, there's data-generating value in taking individual components, blending, and pouring them down my throat to see if anything bad happens. (Julia Galef on tradition as it pertains to social systems, that happens to be applicable here.)

But I'm not very worried; I have trouble imagining a food that has positive effects of "improve body comp, improve mental clarity, clear skin, make semen taste good" and no known negative effects and is biochemically plausible to actually be bad in the long term. Certainly not impossible, but not very probable, I think.

Comment author: michael_b 10 April 2015 01:36:41PM *  1 point [-]

Makes sense, thanks for the link and your summary.

I've taken a keen interest in soylent but am happy to let others beta test long-term effects for me before I give it a shot :)

FWIW, the way soylent people describe their results is more or less how I describe what happened to me when I adopted a whole food plant-based diet (the "china study diet"): BF% dropped/I got leaner, various body odors improved, huge reduction in acne, became a morning person, was able to stop taking ADHD meds, and felt no negative effects at all. Except for maybe I now have so much energy I just had to pick up distance running and ultimately hurt my ankle. :P

Comment author: michael_b 10 April 2015 12:07:54PM *  1 point [-]

The more time I spend hanging out with rationalists the less comfortable I am making predictions about anything. It's kind of becoming a real problem?

"Do you think you'll be hungry later?" "Maybe"

: /

Comment author: zedzed 03 April 2015 08:49:52AM *  13 points [-]

I've been watching for several years now (I adopted the diet myself in 2010), and all of the negative critiques tend to fall into (a) critiques from non-experts, (b) critiques from experts in unrelated fields, (c) health experts who agree that his recommendations have merit, but that they're impractical for the general public to follow.

I produce for you a book written by a relevant expert with ~2.5 times as many references as The China Study (2034 vs 758) who advocates eating an ancestral diet (lean unprocessed meat/fish, fruit, nuts, vegetables/root vegatables) (1). A list of individuals with relevant graduate degrees who more-or-less agree with him can be found in this list of speakers at a paleo conference he spoke at. His recommendations are at least as similar to the recommendations the Mayo clinic returned for me as Campbell's.

That is, I can make a symmetrical argument for a significantly different diet (2), complete with experts and evidence and stuff.

So, to address your questions directly: you should believe that nutrition is a young and complex field, and therefore shouldn't have everything all figured out; my take is that you may do well to replace grains with root vegetables, since that's something everyone agrees is good (plus they're really tasty!); this isn't good enough to inform your dietary choices because I just used a symmetrical argument for a diet that has nonnegligible discrepancies with the diet Campbell recommends; and I don't know how to dig out a signal that experts, to my knowledge, haven't managed to dig out without becoming an expert.

(FWIW, I spent about 5 years as a vegetarian, followed by 1.5 years doing the paleo thing, and now subsist entirely off DIY soylent, which combines the virtues of deriving all its protein from animal sources and being processed.)


(1) Interestingly, Campbell's and Lindeberg's diets can be eaten simultaneously, and this intersection is 100% in-line with what the Mayo clinic recommended me. The difference is that Campbell allows grains and beans, and Lindeberg allows (unprocessed) lean meats, fish, and eggs.

(2) Again, there's substantial overlap, but also substantial disagreement: Lindeberg, for instance, observe the Inuit derive something like 98% of their calories from animal sources and are virtually untouched by Western disease, and concludes that very high consumption of (unprocessed) animals is perfectly fine, whereas Campbell claims that humans should eat minimal amounts of animal.

Comment author: michael_b 10 April 2015 11:23:14AM *  1 point [-]

Separate topic!

(FWIW, I spent about 5 years as a vegetarian, followed by 1.5 years doing the paleo thing, and now subsist entirely off DIY soylent, which combines the virtues of deriving all its protein from animal sources and being processed.)

What I find alarming about soylent-like diets is the idea that you can completely capture human nutritional needs as a table of micronutrients quantities to fill, and then go out and source those individual micronutrients, combine them, and drink.

Aren't you discounting the importance of the configuration of these micronutrients as they arrive in their natural packages? That is, you can certainly decompose an apple into fructose, fiber, vitamins, minerals and water (and etc), but I find it hard to accept that shopping for these individual components, blending, and pouring down your throat is just as good (or better) than eating the apple. Surely we do not completely understand everything nature has done in building us this apple.

Comment author: CronoDAS 05 April 2015 06:51:54PM *  1 point [-]

In U.S. academia, "Professor Emeritus" is a title given to professors who are retired.

Comment author: michael_b 10 April 2015 07:23:36AM 0 points [-]

D'oh! I thought it meant some kind of special honor. Does it at least mean "was granted tenure and was not fired"? That's not useless information, I guess.

Comment author: drethelin 05 April 2015 08:25:13PM 1 point [-]

Why wouldn't nutrition scientists studying nutrition come to a similar conclusion about how young, murky, and complicated nutrition is? Shouldn't they on average know this better than anyone and only make very careful and strongly supported recommendations?

They have very strong incentives (ie earning money and building a career and having patients) to pretend to be certain. People don't want to pay for honest but vague guesses.

If you can't trust nutrition scientists to judge the literature properly, why should you trust scientists outside of the field or layman attempting to dive into the field would be better?

You shouldn't really trust scientists outside the field to talk about the entire field of nutrition but insofar as experts in older and more reliable fields like chemistry or biology disagree with specific nutritional claims you should probably agree with the actual scientists.

Comment author: michael_b 10 April 2015 07:14:43AM *  0 points [-]

They have very strong incentives (ie earning money and building a career and having patients) to pretend to be certain. People don't want to pay for honest but vague guesses.

I would expect consensus (or the lack thereof) is an important signaler for exposing this kind of bias?

Comment author: zedzed 05 April 2015 12:05:46PM 1 point [-]

Lindeberg is a nutrition researcher (conducts studies, co-authors papers) coming from a medical background, which makes him just as much an expert as a nutrition researcher coming from a biochemistry background.

Why wouldn't nutrition scientists studying nutrition come to a similar conclusion about how young, murky, and complicated nutrition is and only make very conservative, very strongly supported recommendations?

We can measure how much a field has progressed by its predictive power, and nutrition is already making concrete predictions with high confidence. Not a lot, not with the confidence of, say, Newtonian mechanics but, given how very much literature there is and how very complicated things are, the level of consensus across researchers who are coming at the problem from disparate-but-legitimate approaches (e.g. biochemical, evolutionary) is sufficiently impressive that I do trust them to judge the literature properly. Humans are biased, so it's unsurprising that we don't yet have a consensus as broad as, say, existence of the golgi apparatus, but the world looks exactly as we'd expect it if nutrition scientists were doing good work in a complicated field.

To summarize: Lindeberg, like Campbell, is an experienced nutrition researcher with impressive and relevant credentials. Nutrition is a young and complex field, so there's no broad consensus about everything—although there is broad consensus about some things—but nutrition scientists are doing a decent enough job of figuring things out that I trust them to judge the literature properly.

Comment author: michael_b 10 April 2015 07:10:04AM *  0 points [-]

Lindeberg is a nutrition researcher (conducts studies, co-authors papers) coming from a medical background, which makes him just as much an expert as a nutrition researcher coming from a biochemistry background

Am I asking for too much by insisting on a nutrition researcher from a biochemistry background to refute Campbell? Or are you saying they can both be right within the framework of their fields?

To summarize: Lindeberg, like Campbell, is an experienced nutrition researcher with impressive and relevant credentials. Nutrition is a young and complex field, so there's no broad consensus about everything—although there is broad consensus about some things—but nutrition scientists are doing a decent enough job of figuring things out that I trust them to judge the literature properly.

I am moved enough by your insight and your persistence to give Lindeberg's book a read. :)

Comment author: ChristianKl 06 April 2015 02:21:30PM 0 points [-]

The China study found some probably correct things assuming they followed protocol and did statistics well

I don't think they did statistics well:

Notice Campbell cites a chain of three variables: Cancer associates with cholesterol, cholesterol associates with animal protein, and therefore we infer that animal protein associates with cancer. Or from another angle: Cancer associates with cholesterol, cholesterol negatively associates with plant protein, and therefore we infer plant protein protects against cancer.

But when we actually track down the direct correlation between animal protein and cancer, there is no statistically significant positive trend. None. Looking directly at animal protein intake, we have the following correlations with cancers:

Comment author: michael_b 10 April 2015 06:51:16AM *  0 points [-]

I've come across this quite often. This is written by an amateur whose authority stems from "I typically spend about five hours a day reading and writing about nutrition—voluntarily".

As a layperson myself I'd be a lot more moved if other nutrition scientists agreed with her. As it stands for me her input is basically +1 "non-nutrition scientists disagree' with Campbell".

Comment author: Ishaan 06 April 2015 07:00:08AM *  -1 points [-]

Based on what I know about the words "professor" and "emeritus" and "cornell", I assume this is written by an authority in the field of nutrition.

The value of being an authority in a field is that you can accurately convey the consensus within that field. Whenever consensus within a field does not exist, the ancient injunction against "argument from authority" remains true. The "authority" derives not from the authoritative individuals themselves, but the collective wisdom of the field to which they've been exposed.

  1. Science is crap. Don't believe expert predictions about the natural world.

Science isn't crap - it's just that when science is weird and inconsistent and wrong it's obvious and everyone notices because science has better epistemic hygiene practices than non-science methods of discovering things. You're just overestimating the degree of accuracy and agreement science aught to have.

I might have the specifics of this story wrong, but once upon a time, scientists (correctly) showed that blood cholesterol is correlated with bad things. So people stopped eating more than 1-2 eggs a day. Now we know dietary cholesterol doesn't directly control blood cholesterol and you can eat lots of eggs and your LDL won't rise in a meaningful way. That doesn't make the original findings wrong, it just means the resulting interpretation was wrong. The original scientists weren't being dumb, it was perfectly reasonable interpretation to make.

The China study found some probably correct things assuming they followed protocol and did statistics well, and now they're interpreting it to mean "go vegan". As a consumer of primary research, you should ask yourself if the findings -> practical interpretation link is reasonable.

nutrition science is crap

It's really not significantly different from "we thought there was Aether, and now we don't", but in fields like economics and psychology and nutrition people notice more when you mess up, because laymen understand enough to know what it means to make a mistake - Everybody votes, deals with minds, and eats, but most people don't understand the implications of their not being an aether. .

But, yeah - scientists are people, and some fields tend get more contaminated with personal biases that the researchers might have acquired entirely outside the laboratory. This can inform the types of questions they study, confirmation bias, bias in the interpretation of the data, and so on. Also, whenever an issue of public policy is at stake, I imagine special interests groups get involved.

Why isn't this good enough to inform your dietary choices?

The evidence certainly does inform my hypothesis, but that's not the same thing as agreeing with the author's interpretation. (I don't say "choices", because my interest is mostly academic and I'm not particularly diet conscious in my personal life, and I do factor in moral concerns with respect to meat, which makes my actual diet not particularly in line with what I think is nutritionally optimal.)

The important thing is to inform ones views based on the evidence gathered, rather than trusting the researcher to interpret their own evidence.

I accept that chinese diets are likely superior to Western diet. Leafs, shoots, roots, fruits, nuts, and the like are all extremely, extremely important. When meat replaces or otherwise funges against those foods (as you'd expect it to given limited calories per day) meat is bad. The average American lives on meat and grains, skipping the fruits and veggies, and that's no good. The average American vegetarian will probably be healthier than the average American omnivore for this reason, even controlling for caloric intake. The average American would probably benefit from going vegetarian, not to mention the various moral horrors and environmental damage meat entails. It's not surprising that we see the same trend in China.

But, well - none of that means that vegan is nutritionally optimal. Hunter gatherers, lacking grains and dairy to provide calories, would have probably "maxed out" the benefits leafs, shoots, roots, fruits, nuts, and other plants despite also eating lots of meat. The Inuit pretty much just eat meat and do fine. (Don't try it at home - the Inuit can only do this through judicious consumption of organ meats, which are glycogen rich and nutrient dense. They actually often discard the lean muscle meat, probably because they've intuitively grasped the macro-nutrient ratio problem it would pose.)

When you look at the evidence provided from the China study - not the interpretations, just the evidence - there's very little room for suggesting that hunter-gatherer diets are suboptimal.

Assuming you don't plan to become an expert in the field of nutrition yourself, what's a better way to inform your dietary choices?

There's no way out. I can say "Eat a balanced diet, with natural real foods" or whatever but the true meaning of "balanced" and "natural" is a lot more controversial than it seems at first. If you don't do your research you are down to guesswork, and after you do your research you are still mostly down to guesswork due to how little we really know.

FWIW, having a totally optimal diet is probably not extremely important in the grand scheme of things. For all practical purposes you'll be just fine so long as (1) your calories are okay, (2) your macronutrient ratios are not horribly awful, (3) you have no obvious symptoms of micronutrient deficiency, and (4) you exercise. Everyone agrees on those four and that you should eat fruits and veggies. Beyond that, it's all controversial and I doubt the additional worrying will really buy you that many extra years of quality health when one factors in the likelihood of choosing the correct arguments among the controversy. (Outside view. Inside view, I totally think I'm right in choosing the "take cues from hunter gatherers" thing.)

Comment author: michael_b 10 April 2015 06:43:35AM *  0 points [-]

Based on what I know about the words "professor" and "emeritus" and "cornell", I assume this is written by an authority in the field of nutrition.

The value of being an authority in a field is that you can accurately convey the consensus within that field. Whenever consensus within a field does not exist, the ancient injunction against "argument from authority" remains true. The "authority" derives not from the authoritative individuals themselves, but the collective wisdom of the field to which they've been exposed.

Who says there is no consensus? Given that he's a nutrition science authority, and that other nutrition science authorities aren't refuting him, that's some small evidence that he's representing a consensus (there are other possible explanations as well, that I touched on in my OP).

View more: Next