Evidence against Calorie Restriction

post by Vaniver · 2011-11-16T18:08:38.397Z · LW · GW · Legacy · 23 comments

An article about the pitfalls of using mice for animal research leads off with the example of calorie restriction. The controls for calorie restriction mice experiments were obese mice, suggesting that the health benefits of calorie restriction might be conflated with the health benefits of not being obese. I get the impression that people who study calorie restriction still think it worthwhile for life extension, but it's useful to try and integrate all evidence you come across.


[edit] timtyler suggests this is a well-understood effect that's already been taken into account by CR scientists.


Comments sorted by top scores.

comment by timtyler · 2011-11-16T19:02:21.374Z · LW(p) · GW(p)

It is standard practice - by those in the know - to use restricted animals as controls in CR experiments - in order to help to avoid this criticism. The controls actually need to live longer than normal, to help prove that the researchers haven't screwed up somewhere with diet, conditions, pathogens - or whetever.

Some researchers don't do this - and their evidence is promptly discarded by those who understand the issue.

Anyway, it hardly seems fair to bill this effect as "Evidence against Calorie Restriction". This is a well-known and well understood effect for those in the field. Dietary energy restriction enthusiasts are not basing their diet on this kind of screwed-up evidence in the first place.

Replies from: JoshuaZ, Vaniver
comment by JoshuaZ · 2011-11-16T19:08:45.335Z · LW(p) · GW(p)

It is standard practice - by those in the know - to use restricted animals as controls in CR experiments - in order to help to avoid this criticism.

Can you expand this statement? I'm not sure what you mean. You seem to be distinguishing between "restricted" and "caloric restricted" but it isn't obvious what distinction you are making.

Replies from: timtyler
comment by timtyler · 2011-11-16T19:28:50.906Z · LW(p) · GW(p)

No such distinction was intended.

"Dietary energy restriction", "calorie restriction", "CR" and "restricted" were intended as synonyms.

Replies from: JoshuaZ
comment by JoshuaZ · 2011-11-16T19:34:40.541Z · LW(p) · GW(p)

Then what do you mean by:

It is standard practice - by those in the know - to use restricted animals as controls in CR experiments

One is comparing CR animals to non-CR animals. What do you mean by this?

Replies from: timtyler
comment by timtyler · 2011-11-16T19:41:04.337Z · LW(p) · GW(p)

I don't understand your question. "Proper" CR experiements use skinny experimental animals and skinny controls. Maybe 20% and 40% restricted - precisely to make sure that the control animals are not sick and obese.

Comparing with obese animals is bad - for all the reasons recently elaborated on - however this is really very old news in the community that are actually on such diets - and among most of the scientists involved.

Replies from: ahartell, JoshuaZ
comment by ahartell · 2011-11-16T19:47:39.608Z · LW(p) · GW(p)

I understand what you're saying after reading this comment, but the quotation below still seems unclear.

It is standard practice - by those in the know - to use restricted animals as controls in CR experiments

"Dietary energy restriction", "calorie restriction", "CR" and "restricted" were intended as synonyms.

With no distinction between restricted and calorie restricted, it seems like you are saying that in Calorie Restriction experiments the control group mice are on calorie restricted diets, as presumably are the test group mice. This is confusing to me.

I understand now (the control group mice are restricted but less so), but maybe the first comment could be edited for better understanding?

comment by JoshuaZ · 2011-11-16T19:47:39.402Z · LW(p) · GW(p)

Ah ok. I got the point. The point is that one uses different degrees of restriction and doesn't include any animals that could be classified as obese.

comment by Vaniver · 2011-11-16T20:02:02.292Z · LW(p) · GW(p)

This is a well-known and well understood effect for those in the field.

Good! I was hoping this would be the case.

comment by Douglas_Knight · 2011-11-16T20:50:50.149Z · LW(p) · GW(p)

The combination of the example you gave, CR, and the subtitle of the post you link to "The dangers of using one lab animal to study every disease." is 100% bullshit: CR has been studied in a huge number of animals, from worms to primates. I foolishly read a few paragraphs, and it sounds like Mark Mattson simply hasn't read the literature on CR.

Replies from: timtyler
comment by timtyler · 2011-11-17T01:34:47.364Z · LW(p) · GW(p)

Nope: "Mattson is among the leading researchers on the effects of calorie restriction and the brain." - source.

comment by Manfred · 2011-11-16T22:28:30.972Z · LW(p) · GW(p)

Also, skinny people die more than people near the border between normal and "overweight," for reasons not fully understood, though which may be at least partially due to tobacco use.

Replies from: timtyler, NancyLebovitz, Vladimir_Nesov
comment by timtyler · 2011-11-17T01:36:28.212Z · LW(p) · GW(p)

Being sick often makes you skinny. There are many causes of muscle wasting disorders, including many kinds of malnutrition.

Replies from: Manfred
comment by Manfred · 2011-11-17T04:19:44.010Z · LW(p) · GW(p)

Data collection on deaths began 5 years after collecting the BMI to try to limit this sort of effect, as they state in the summary. And they do a cause of death breakdown which eliminates the idea of a muscle-wasting disease.

Replies from: timtyler
comment by timtyler · 2011-11-17T13:05:34.358Z · LW(p) · GW(p)

60% of the excess mortality at low BMIs in the study was down to respiratory disease, most of which is due to smoking. Being skinny is a marker for being sick and for being old - both of which tend to come on before death, but neither of which are necessarily confined to the last 5 years of life. This kind of statistical association between being skinny and being ill is thought to explain most of the effect of increased mortility at low BMIs.

The question calorie restriction practitioners want the answer to is: "what is the effect of low BMI caused by low intake of dietary enegry while ensuring good nutrition on lifespan?" There is data on this question, but it mostly doesn't come from studies like this one - where the signal is dominated by people who are skinny because they are sick, old, malnourished or had developmental issues.

A good number of the experts who have considered this issue seem to think positive efefcts on health and lifespan in initially healthy western humans on sensible energy-restricted diets are likely to go down at least to to BMIs around 18. Some take their own weights down below that - but mostly confess that they are exploring rather poorly-charted territory.

comment by NancyLebovitz · 2011-11-17T14:22:48.925Z · LW(p) · GW(p)

I've been going with the hypothesis that people with a little more fat that what's considered healthy have more reserves for illness, operations, and/or cancer treatments.

I believe tobacco use has been considered in at least some of those studies.

So far as CR is concerned, a lot of thin people (like anyone else) aren't getting optimal nutrition.

comment by Vladimir_Nesov · 2011-11-16T22:34:11.050Z · LW(p) · GW(p)

(When I lightly researched a while ago one of the topics the article you cited considers (correlation of health with weight in normal to overweight interval), I found that there is no consensus and there are studies giving contradictory conclusions. For causation, it's even less clear, and the current position seems to be that weight doesn't have a known causal role if around normal/overweight, instead exercise and diet affect both weight and health.)

comment by gwern · 2011-11-16T21:14:01.030Z · LW(p) · GW(p)

There is a similar confound for intermittent fasting: if the IF animals lose weight, then that strongly implies that they were not doubling up on their feeding or whatever they needed to do to eat a normal amount of calories. Hence, weight loss + longevity increase ceases to be evidence for IF, and becomes evidence for CR.

Replies from: timtyler, Mercurial
comment by timtyler · 2011-11-17T13:17:51.847Z · LW(p) · GW(p)

There is a similar confound for intermittent fasting: if the IF animals lose weight, then that strongly implies that they were not doubling up on their feeding or whatever they needed to do to eat a normal amount of calories.

Some intermittent fasting studies have controlled for calorie intake, though. For example, this one:

Intermittent fasting dissociates beneficial effects of dietary restriction on glucose metabolism and neuronal resistance to injury from calorie intake.

It doesn't entirely rule out the possibility of reduced calorie intake being responsible - since consumed calories may not result is absorbed calories so reliably on intermittent fasting regimes. However, most explanations seem to focus on the more persistently lowered blood glucose levels and/or the increased stress.

comment by Mercurial · 2011-11-17T03:55:15.747Z · LW(p) · GW(p)

Could you clarify this just a bit? I practice IF for health reasons, so I'm quite interested in learning if it's bunk. I was under the impression that the whole point was that in induces essentially the same effects as CR but without needing to worry about calorie-counting, etc. It's just automatic. Is that what you meant to say? Or are you pointing toward evidence against IF such that people like me who are interested in longevity should consider stopping?

Replies from: gwern
comment by gwern · 2011-11-17T15:25:47.325Z · LW(p) · GW(p)


Replies from: Mercurial
comment by Mercurial · 2011-11-17T17:13:55.302Z · LW(p) · GW(p)

Ah! Thank you. So the short version is:

  • IF is a riskier method of doing CR
  • You still have to have the net effect of CR

My personal experience with IF suggests that there are pretty major benefits from doing it regularly. I lost 30 pounds (and I had no idea I needed to lose them - but I definitely look and feel much better!), my general levels of inflammation dropped, my energy stabilized, and my various biomarkers went from good to excellent. And I'm no longer trapped by a regular need for food; if it takes an extra three hours or so to get to a meal, it doesn't really affect me.

But I've seen this warning about sudden adult-onset CR shortening lifespan, and I see in this article you link to that it's apparently an issue with IF too. That concerns me. I switched to IF over the course of about a week. Does anyone know what the mechanism for shortening lifespan with too quick an adaptation to CR is? If there's some way of checking for that kind of damage or ongoing effect in myself, I'd like to do so.

And is it a matter of "once done, you've done the damage and it's too late," or is it more that ongoing CR that started too suddenly keeps affecting the organism in negative ways, kind of like getting continually exposed to cold air actually inhibits circulation rather than strengthening it if the initial burst of cold was too strong? That difference would matter a lot to me because if IF isn't really having any negative longevity effects, I'd still want to keep doing it because the other benefits I'm getting from it are so pragmatically awesome. But if I'm continually cutting months off my life, I'll stop right away!

Replies from: gwern
comment by gwern · 2011-11-17T17:51:24.844Z · LW(p) · GW(p)

I'm afraid I don't really know. Intermittent fasting, like masturbation, is something I'm still collecting citations and quotes on because no clear picture has emerged for me.

Replies from: Mercurial
comment by Mercurial · 2011-11-17T18:46:58.180Z · LW(p) · GW(p)

Drat. Well, do keep me posted, and I'll keep an eye out for similar info.