Thursday, January 03, 2013

On Fat And Mortality. The Recent Meta-Analysis.


Fat.  It's one of the great fears of today, both in the foods we eat and on our bodies.  Even talking about fat always creates a debate.  And the reasons are in the odd mixture of concepts we have nailed to the word "fat." 

Fat is bad for you, goes the recent medical wisdom.  Yet our bodies need some amount of fat to function, and fat babies were viewed with great admiration only a hundred years ago or so because when starvation is a real possibility having a bit of extra fat is very useful.  Similarly, what was regarded as normal weight, in the terms of "beautiful",  has changed over time.

What makes the debates about fat so very nasty is the moral, even prudish tone.  Being fat is seen as a behavioral problem, as a form of moral failure, as one of those deathly sins:  greed or gluttony, made manifest.  It's one of the human vices one cannot hide the same way one can hide, say, cruelty or avarice.  It's viewed as ugly.  Fat people have no willpower!  Fat people are greedy!  Fat people are Lesser Than Us Thin People.

All that is over and above any medical arguments about overweight or obesity.  It's the moralizing zeal of others which truly hurts anyone labeled overweight and the odd assumption that one can make those moralizing comments openly because, after all, being fat is bad for you. 

It's that moralizing aspect which always leaves me with a sour taste in my mouth because it reveals so very much about those doing the moralizing.  It reveals something about them which should be included in the seven deadly sins or similar lists:  Lack of empathy, perhaps, or the idea that one doesn't have to try to understand individual differences, different life challenges (poverty, say), different metabolic rates or different genetic inheritances; one can simply attribute a reason for someone else's overweight, and that reason is a moral failure.  The assumption beneath that is that everyone is exactly the same in those aspects, that everyone fatter than the judge must have just failed where the judge would have succeeded.

Yet we don't know that.

Because of that moral and judgmental jungle I always find writing about the possible medical aspects of obesity difficult.  You press one of the medical buttons by, say, discussing the correlation between obesity and Type Two diabetes, and you also open one more avenue for the various kinds of ranking judgments.  Those are the reasons why the research about obesity and health should be scrutinized very carefully.  It has consequences over and beyond any health consequences because we use fat as one of the measuring sticks in our current society. 

All that is a long introduction to a new meta-analysis which looks at the relationship between one often-used measure of possible obesity, the Body Mass Index (BMI) and mortality rates.  The meta-analysis was discussed in the New York Times:

The report on nearly three million people found that those whose B.M.I. ranked them as overweight had less risk of dying than people of normal weight. And while obese people had a greater mortality risk over all, those at the lowest obesity level (B.M.I. of 30 to 34.9) were not more likely to die than normal-weight people.
The report, although not the first to suggest this relationship between B.M.I. and mortality, is by far the largest and most carefully done, analyzing nearly 100 studies, experts said.
But don’t scrap those New Year’s weight-loss resolutions and start gorging on fried Belgian waffles or triple cheeseburgers.
Experts not involved in the research said it suggested that overweight people need not panic unless they have other indicators of poor health and that depending on where fat is in the body, it might be protective or even nutritional for older or sicker people. But over all, piling on pounds and becoming more than slightly obese remains dangerous.
“We wouldn’t want people to think, ‘Well, I can take a pass and gain more weight,'” said Dr. George Blackburn, associate director of Harvard Medical School’s nutrition division.
Rather, he and others said, the report, in The Journal of the American Medical Association, suggests that B.M.I., a ratio of height to weight, should not be the only indicator of healthy weight.
“Body mass index is an imperfect measure of the risk of mortality,” and factors like blood pressure, cholesterol and blood sugar must be considered, said Dr. Samuel Klein, director of the Center for Human Nutrition at Washington University School of Medicine in St. Louis.
Dr. Steven Heymsfield, executive director of the Pennington Biomedical Research Center in Louisiana, who wrote an editorial accompanying the study, said that for overweight people, if indicators like cholesterol “are in the abnormal range, then that weight is affecting you,” but that if indicators are normal, there’s no reason to “go on a crash diet.”
Experts also said the data suggested that the definition of "normal" B.M.I., 18.5 to 24.9, should be revised, excluding its lowest weights, which might be too thin.
The study did show that the two highest obesity categories (B.M.I. of 35 and up) are at high risk. “Once you have higher obesity, the fat’s in the fire,” Dr. Blackburn said.

I haven't read the actual report which makes it hard to state how good the meta-analysis might have been.  I hope that they have taken into account the obvious problem with correlations between weight and mortality:  In many terminal diseases one loses weight, and thus BMI, before death.  That fact could appear as a false correlation between low values of BMI and early death.  What we want to study are levels of weight over a longer period of time, of course.

The second potential problem with this study is the concept of BMI as a measure of adiposity.  It makes no real difference between fat and muscle, for instance, and I don't think it allows for different skeletal structures in individuals of the same height. 

But more crucially, crude measures such as BMI don't necessarily tell us what the "normal" BMI should be for individuals of different ages.  It's possible that getting a bit heavier as one ages is, indeed, "normal," in the sense of not being necessarily bad for one's health. 

And we don't really know what the "normal" BMI should be, in general, because how to define that seems to be partly determined by the health consequences of various BMIs.  Hence the suggestion in this study that the lowest BMIs should be removed from the definition of "normal."

Note the circularity in this?  Another reason to be careful when interpreting medical findings of various types.

The NYT write-up also points out that where the body fat is deposited might make a difference in its possibly impact of health.  Stomach fat seems to be the current rogue, whereas pear-shaped people might be OK.   On the other hand, I have read at least one study which argued that greater stomach fat after menopause could benefit women by producing more estrogen which protects against heart disease.

What is the significance of this particular meta-analysis?  It's one analysis of existing studies.  As such, it doesn't offer us brand new evidence.  But whatever its problems might be or might not be, the debate it created tells me, at least, how very invested we humans are in anything regarded as received wisdom.  Of course that received wisdom may well turn out to be correct, in the long run.  Or perhaps not.

To put all this into some kind of perspective, have a look at one proposal in the UK:

Obese and other unhealthy people could be monitored to check whether they are taking exercise and have their benefits cut if they fail to do so under proposals published on Thursday by a Conservative-run council and a local government thinktank.
Westminster council and the Local Government Information Unit say new technologies such as smart cards could be used to track claimants' use of leisure centres, allowing local authorities to dock housing and council benefit payments from those who refuse to carry out exercise prescribed by their GP.

That proposal is based on the assumption that obese people, however obesity might be defined there, are unhealthy without any other evidence of that, and therefore obese people could be monitored and punished if they failed to carry out physician-prescribed physical exercise.  Non-obese people would not be monitored unless they, also, were defined as "unhealthy." 

Suppose that we decided, on the basis of that recent meta-analysis, that people with very low BMIs are now "unhealthy."  Should we institute structures of punishment for them, too?  Some sort of a smart card which measures their calorie consumption, with docking of benefits if it is too low?

I find that proposal extremely distasteful unless all British patients under similar conditions are monitored and punished for failed compliance in the treatment of their conditions.  That this is suggested for obese people is probably because of that moralizing aspect about obesity.

I should stress, to wrap up this post, that I'm not insisting that it doesn't matter what people weigh or that the medical studies shouldn't continue to analyze the possible harmful effects of overweight and underweight and so on.  But I think we have jumped too far to the other direction in our moral and ethical treatment of obesity and overweight as a sin, rather than something worth studying in the medical field.  It's salutary to discuss that aspect of the so-called obesity epidemic, at least.