Greta Christina's fat positive manifesto

[ciphergoth]

I froze the discussion here because I thought it deserved a top-level post of its own, rather than being under a general discussion of Greta Christina. A few weeks ago she posted a very interesting series of articles on the fat-positive movement and her own beliefs; I'd be very interested to read more about what people think of them.

"I was frankly shocked at how callous most of the fat-positive advocates were about my bad knee. I was shocked at how quick they were to ignore or dismiss it. They were passionately concerned about the quality of life I might lose if I counted calories or stopped eating chocolate bars every day. But when it came to the quality of life I might lose if I could no longer dance, climb hills, climb stairs, take long walks, walk at all? Eh. Whatever. I should try exercise or physical therapy or something. Oh, I'd tried those things already? Well, whatever."

• The Fat-Positive Diet
• The Fat-Positive Skeptic
• An Open Letter to the Fat-Positive Movement

Comments

[wildeabandon.livejournal.com]

I'm largely in agreement. The vast majority of what the fat-positive movement stands for should be uncontroversial, but to quote your response to the third post:

"I agree that unless you are an expert - and by expert, I mean on the level where your research is published in peer-reviewed journals - you are crazy to bet against the scientific consensus."

I haven't done enough reading of the literature to be confident in what the scientific consensus is, and to what extent it is scare-stories in the media, but I've certainly seen more papers arguing that there is a negative correlation link between weight (above a certain level, obviously) and a variety of health measures than I've seen arguing the opposite.

[ciphergoth.livejournal.com]

I'd be willing to take a bet on what answer I'd get if we asked the head of a randomly chosen University medical department - and if I'm wrong about this I'd like to know very much.

[valkyriekaren.livejournal.com]

Certainly the learning objectives here at St George's are keyed towards an understanding of BMI and the importance of lifestyle changes (including activity level, dietary restrictions and weight loss as appropriate) in controlling conditions such as high blood pressure, high cholesterol and heart disease.

(I still think BMI is unreliable and shouldn't be the be-all and end-all, and that the lines for 'healthy' and 'overweight' are drawn in completely the wrong places. But that doesn't mean it's not an OK place to start when looking at overall physical health.)

[ladycat.livejournal.com]

I think maybe bodystat is a beter measure than BMI because it gives you a clearer overall picture of the composition of your body.

[valkyriekaren.livejournal.com]

Indeed, but at the moment not all GP surgeries, hospitals, clinics and health centres have bodystat machines (which are quite new and expensive), whereas they do all have scales and height charts (old and cheap).

[wildeabandon.livejournal.com]

One thing that I have a suspicion is supported by the science, but is a long way from permeating through to the popular consensus, including medics, is the fact that diets usually don't work.

[purplerabbits.livejournal.com]

I think some of it has permeated, but most individual GPs seem to feel that the *reason* most diets don't work is that eithyer they are crash diets or the dieters aren't doing it right. Hence their tendency to just harangue people harder into following what they consider to be a sensible diet.

What I think they fail to take into account is the cognitive bias which can happily accept that 90% of something is crap and just assume that the thinker falls into the 10%, plus that even if they're right, pushing people harder also doesn't work.

[http://users.livejournal.com/_lj_sucks_/]

"Diets don't work" is an oversimplification. Diets work fine if they are permanent lifestyle changes, i.e. permanently changing you diet. What doesn't work is dieting to lose weight and then stopping that diet and going back to what you were doing before. Which ought to be so obvious that it doesn't need saying, but clearly isn't obvious to customers of the diet industry...

[wildeabandon.livejournal.com]

Oh, I think these days a lot of the diet industry talks about permanent lifestyle changes. But they generally don't work either, because they're very difficult to stick to. If they weren't difficult to stick to, then failure rates wouldn't be so high

[lizw.livejournal.com]

In this particular case, I think there are strong reasons to bet against the scientific consensus, starting with the fact that almost all of the science in question is of appallingly poor quality and skewed by a combination of politics and big money. I'd recommend picking out some specific studies and looking to see whether junkfoodscience has commented on them, or reading her obesity paradox series.

[ciphergoth.livejournal.com]

I'd recommend picking out some specific studies and looking to see whether [info]junkfoodscience has commented on them

This is a good recommendation. I can't find anything on any of the studies referenced here on junkfoodscience, though that's not to say they're not discussed.

Edit: I note they have a post specifically about junkfoodscience.

Edit: and she links with approval to global warming denialist Steven Milloy's junkscience.com. Top link on her list of recommended links. Last is noted nut Bjorn Lomborg. Um.

[valkyriekaren.livejournal.com]

Edit: and she links with approval to global warming denialist Steven Milloy's junkscience.com. Top link on her list of recommended links. Last is noted nut Bjorn Lomborg. Um.

I'm not sure damning someone for the company they keep is much of an argument.

[ciphergoth.livejournal.com]

I agree, but I think it's informative all the same.

[blackmetalbaz.livejournal.com]

It's amazing how many cranks there are out there that do ascribe to multiple branches of pseudoscientific thinking (of which this is one). I think that in itself can be very telling.

[djm4]

OK, maybe I do have a little time...

I just lost about an hour following the cited links on the scienceblogs piece. My conclusion: evidence of correlation is not evidence of causation. Even if Mark Hoofnagle says it is. Even if he cites six papers in a row showing correlation, but none showing causation. Even if he wishes very, very hard.

But you know this; you posted the nifty pirates vs global warming graph.

As you say, um.

(Note - please don't take my word for it, but follow the links yourself. You may see something I didn't - and with most, I do only have the abstracts to go on.)

I'm not saying MarkH is wrong, I'm just saying that the evidence he cites doesn't appear to back him up. There may be other evidence that does. He may be reverse-cherry-picking, for some reason known only to himself.

[ciphergoth.livejournal.com]

I don't immediately see how one might run a study which overcame this objection; if you can, could you outline it?

[lizw.livejournal.com]

For anyone who accepts that there is a correlation strong enough to warrant investigation of possible causation, presumably the next step would be to try to design studies to demonstrate the mechanism by which causation occurs. Perhaps not literally the same people, since the specialisms would probably be different, but funding bodies, say. I think one of the problems with this debate is that public health bodies actually jump directly from epidemiological studies that suggest correlation to health recommendations that assume causation, without making it clear that the mechanism is not understood, and therefore the recommendations are provisional at best (I would argue often premature, given that they are often also associated with serious risks.)

[djm4]

Quite apart from what lizw said, I see the value in long-term studies to see whether controlling weight is (a) predictably possible and (b) correlates with health benefits. This is one reason why I feel I need far more time than I have to research this; some such studies appear to have been done and the results are mixed. Diabetes seems to be a better target than blood pressure, and the reporting seems to assume that this is down to the people in question loosing weight even though the study doesn't actually usually (ever?) say so. Now, that might be because it was simply too obvious to mention in the study, but I'd like to see the figures. I may also have been looking at the wrong studies and - as always - getting to anything more than an abstract is rare.

Really, though, I simply haven't done enough research. I know that, and it makes me very uncomfortable. I'm very aware of the parallels between the way fat-positivity is argued and the way, say, climate change denialism is argued. However, there are also parallels with the some of the 'science shows obesity causes diabetes' studies and with the 'science shows that male bisexuality is a myth' study, and much of the language of the fat-acceptance movement is very similar to the language used by Ben Goldacre in Bad Science. Ultimately, whatever my unease about it, I try to read the green ink and see what I think that studies linked to by both sides actually show.

I do know that my experience with that Mark Hoofnagle article you linked to is typical. People say 'this study shows that obesity causes diabetes', when in fact the study shows a correlation. I'm not dismissing correlation as meaningless - possibly fewer pirates did cause global warming; it's certainly worth looking a reason for the correlation, even if in that case the reason is pretty obvious - but I don't think it helps the case of medical professional to mischaracterise it.

I'd also just like to highlight this paragraph:

"Losing weight [in the elderly - this is leading in from a paragraph talking about them rather than the population in general] therefore appears to be a risk for death, and it is also possible that dieting in an older population simply isn't a safe proposition anyway. The message you should take home from all these studies of obesity and weight loss or gain is simple. It is very difficult to improve health through making people lose weight, and diets rarely have long lasting effects. Exercise, even in the overweight, is the most likely intervention to improve markers of cardiovascular health. In the overweight, appropriate management of symptoms like hypertension, diabetes etc., is effective in decreasing mortality. And finally, the best way to prevent the diseases of obesity is to avoid obesity in the first place. It's called primary prevention."

Ignoring, for a moment the last two sentences (because I think it's a separate issue - I'm happy to consider them if you disagree), this is pretty much exactly what most fat-acceptance campaigners are saying, too.

[wildeabandon.livejournal.com]

My conclusion: evidence of correlation is not evidence of causation

But correlation is evidence of either correlation in one direction or the other, or a common cause (or, as I suspect is more likely, some combination of the three). Also, one of the problems with (parts of) the fat positive movement, and Sandy in particular, is that they seem to be denying that there's even correlation.

[lizw.livejournal.com]

I think Sandy's position is that some studies show a weak correlation at a level which is not normally thought sufficient to warrant further investigation of causation.

[djm4]

But correlation is evidence of either correlation in one direction or the other,

You mean 'but correlation is evidence of either causation in one direction or the other...', don't you?

[lizw.livejournal.com]

Yes, I don't trust her on issues other than fat and nutrition; she has some odd political views which skew her judgment outside her immediate area, and of course once you've arrived at the suspicion that Western science has failed colossally in one area, it becomes more tempting to think that it might do so in others. I've followed back enough links from her articles to studies on obesity and read enough around the subject generally to trust her quite a lot on that, though. She's hardly the only person to show good judgment on issues in her field and bad judgment on issues outside it. (And yes, I'm commenting on issues outside my field here myself, and I fully expect you to give my comments less weight on this than I hope you would on, say, refugee law.)

On the one part in your second link where they are actually discussing her reporting of a recent study, I don't see that their commentary really contradicts her. Fat people do better in later life because they can better afford the weight loss that comes with many illnesses of longevity, and when it comes to mortality, that outweighs any adverse health effect of obesity. For whether there are any such adverse health effects that fall short of a net effect on mortality, you need different studies, and I don't think she argues otherwise.

I can't find any discussion on her blog of the New England Journal of Medicine study mentioned in your first link either, but I'd certainly heard of the study. Paul Campos has criticised it here. The paper itself is here.

[booklectica]

That first article doesn't seem to address the idea that it's much better for your health to be rich and fat than poor and thin. (I know you weren't necessarily quoting it because you agreed with it, I just wanted to post that link.)

[palmer1984.livejournal.com]

Don't have time to read the article so sorry if it addresses this, but being poor is generally very bad to your health anyway.

[booklectica]

That's basically what the article is saying, yes. (The one I link to, I mean.)