The Obesity “Myth”

Obesity. The controversy about it rages on. And just like global warming and the Holocaust, it has its share of deniers. …  It doesn’t exist; it is only a false category created by definition. … It may exist, but diets don’t work so there’s nothing to be done about it. … It may exist, but it’s harmless. … It exists, but not only is it harmless, it’s actually healthier. … It exists, but dieting is far more harmful than the extra pounds.

For myself, images like those above drive it home that something unprecedented has happened in those 13 years that has dramatically changed the amount of extra pounds we Americans are hauling around. And it’s continuing. That dark southern underbelly of the US hanging down between those two legs continues to get darker. The 2010 CDC map has 11 states in the south at over 30%.

Even more concerning is the ever increasing diabetes prevalence trailing just a few years behind:

I find it flabbergasting that people can be convinced that all those extra pounds are benign, much less healthy, so I decided to look into it to see if maybe I was wrong and they were right. Here’s just the tip of the iceberg of what I found. (Spoiler: it doesn’t support the obesity deniers.)

The following discussion of intentional weight loss and mortality is taken from the Google books preview of the textbook, “Progress in Obesity Research: 9”, p 264 (yellow highlighting is mine):

Only a few observational cohort studies have examined the association between intentional weight loss and disease outcomes. A series of analyses that used data from the Cancer Prevention Study suggest that the association between intentional weight loss and mortality varies across different population subgroups and by cause of deaths[9-11]. Among a general population of overweight women aged 40-64 years, intentional weight loss was associated with 20 per cent reduced mortality among those with pre-existing conditions but not among healthy women. The reduced mortality among women with health conditions was primarily due to a 30-40 per cent reduction in cancer mortality; there was only a weak (10 per cent), non-significant association with cardiovascular disease (CVD) mortality. In a separate analysis of persons with diabetes. intentional weight loss was associated with 25 per cent reduced all-cause mortality and a 28 per cent lower CVD mortality, with weight loss most protective (33 per cent reduction in all-cause mortality) among those reporting a 20-29 lb weight loss[10]. Among a general population of men, however, intentional weight loss was not associated with mortality regardless of pre-existing health status[11]. Unintentional weight loss in these analyses was either unrelated to mortality (men overall, women with pre-existing conditions) or associated with higher mortality (persons with diabetes, healthy women)[9-11]

Among women aged 55 to 69 with pre-existing health conditions enrolled in the Iowa Women’s Health Study, there was an elevated mortality rate associated with intentional weight loss, but this finding was not statistically significant. Additionally, there was no association between intentional weight loss and mortality in the overall sample but unintentional weight loss was associated with increased mortality[12]. One other study (the Israeli Ischaemic Heart Disease Study) found higher mortality rates for persons who lost weight on a ‘slimming diet’ than for those who were not on a diet, but there was not adjustment for health status, smoking, behaviours, or obesity[13]. Thus, it is difficult to interpret their findings because of the lack of control for self-selection of sicker persons among the dieting group.

 

More recently, an analysis of 9-year mortality follow-up data of persons with diabetes from the 1989 National Health Interview Survey suggested that ‘trying to lose weight’ may be a more important predictor of mortality than actual weight loss[14]. Persons who were trying to lose weight had a 23 per cent lower mortality rate (HRR 0.77, 95 per cent Cl. 0.61—0.99) irrespective of weight change. Compared to persons reporting no weight loss and not trying to lose weight, persons who tried to lose weight but failed had a lower mortality rate (RR 0.73, 0.55 to 0.96) while lose who succeeded (i.e. intentional weight loss) had a non-significant reduction in mortality (RR 0.82, 0.62—1.08). Persons with unintentional weight loss had a 58 per cent increased all-cause mortality rate.

 

Taken as a whole, studies have not consistently supported an association between intentional weight loss and mortality, and with the exception of the study conducted among persons with diabetes by Williamson et al.[10], there is no evidence for a dose-response association. These studies appear to refute the notion, however, that intentional weight loss is harmful[8]. At least three limitations should be considered when examining this literature. (1) the associations between weight loss and mortality reduction may be underestimated due to error in self-reported weight change; (2) the groups reporting intentional weight loss may be inflated by persons simultaneously undergoing unintentional weight loss related to underlying disease: if so, this would lead to an underestimate of the benefit of intentional weight loss; and finally; (3) there may be unexplored but important associations with disease incidence and quality of life outcomes that are not detected by studies of mortality.

 

But why was I looking at this $200 book? Well, Paul Campos wrote this in “The Obesity Myth”:

… the Iowa Women’s Health Study – produced some rather extraordinary data in regard to the assumption that trying to get thin is the appropriate ‘cure’ for the ‘disease’ of above-average weight.  The Iowa study is particularly striking, in that it featured no less than 108 different statistical comparisons, based on age, initial weight and health status, and cause of death.  In seventy-nine of these comparisons, intentional weight loss was associated with higher mortality rates.  By contrast, the number of comparisons in which intentional weight loss ended up being associated with lower mortality rates was zero.

which was subsequently quoted extensively by Matt Stone in his “12 Paleo Myths” book where I came across it. Mr. Stone was using the quotes in an attempt to prove that overweight and mild obesity is at worst harmless and might even be good for us. Even though there was no mention in Stone’s book about the controversial nature of the material he was presenting, it raised huge red flags for me. So Stone’s book was the catalyst that sent me off to pubmed and the obesity textbook was the clincher that told me after a couple of weeks of reading that I finally knew enough about the obesity epidemic as “myth” to lay it to rest, as in R.I.P.

First off, I gotta say it was a bit disconcerting and mindboggling to me to discover that the authors of the obesity textbook were reporting a diametrically opposed interpretation of the exact same data regarding effect of intentional weight loss on mortality from that of Paul Campos. But let’s see if we can make sense of this.

So who is Paul Campos? From wiki:

 Paul F. Campos is a law professor, author and blogger on the faculty of the University of Colorado Boulder in Boulder. Campos received his A.B. (1982) and M.A. in English (1983) from the University of Michigan and in 1989 his J.D. from the University of Michigan Law School.[1] Campos worked at the law firm Latham & Watkins in Chicago from 1989-1990 and became an Associate Professor at the University of Colorado in 1990, where he teaches classes on property, punishment theory, jurisprudence, and legal interpretation.[2] 

Campos’s contentions that obesity is healthy were … overwhelmingly criticized by medical, epidemiological, and statistical researchers with professional training in empirical research.[4][5][6][7]

Sadly, Campos has been identified

… as one of three scholars (along with University of Chicago political scientist Eric Oliver and Arizona State University exercise physiologist Glenn Gaesser) who have led the way in constructing a counter-narrative to the claims of those who argue higher body weight represents a major public health crisis

and criticized for “overstating the empirical case for obesity skepticism, and for relying too much on criticisms of the motives of certain obesity researchers.”

So what does it mean when Campos finds 79 out of 108 different statistical comparisons in the IWHS supporting his premise? Well, overlooking the contention that he might be cherry-picking his data by focusing on the least strong of the obesity studies, what it means is: nada, zilch, zippo! In fact, the most forgiving and generous explanation is that what Campos probably doesn’t know (because he studied law and not statistics) is, first of all, the most important feature in statistics is what’s called the P-value, ie the probability that the result you are seeing is due to chance. You want your P-values to be low. And here’s the rub. The more associations you look at, the greater is the chance that random sampling noise will produce something that looks significant. But Campos doesn’t bother to compute a P-value for his conclusion. And secondly, what we also learn in preparation for a career in medical research is that a single study means nothing. We must evaluate the body of research as a whole. So, have a look again at all the research I highlighted from the obesity textbook, not only the IWHS which failed to achieve significance.

I can guess what Campos would think about me as a mathematician trying to teach a course in law. So why is Campos trying to teach these epidemiologists how to interpret study results?

But a more interesting question is, how can statistics render an entirely different interpretation than so-called “common-sense”? Doesn’t 79 out of 108 win the day? In a word, no! Suppose the people that choose to diet are a little older or a little more ill (and thus more concerned with their health) than those that don’t. They will have a higher death rate simply because when it comes to dying, age and illness trumps everything. The only appropriate yardstick is to use is to compare a group’s actual death rate against it’s expected death rate when adjusted for all the other factors that affect mortality. It’s called multivariate analysis. Campos didn’t do it and, thus, shouldn’t even be opening his mouth.

So why do I find this so sad? Look again at the Campos book that the quote is from:

Right there in the title is the problem. Campos claims it’s hazardous to your health. But if you look again at the line I underlined above in the quote from the obesity research book,

“These studies appear to refute the notion, however, that intentional weight loss is harmful.”

Of course, it’s comforting to hear that your weight problem isn’t really a problem and pass the pizza and beer, please. It’s what we really want to hear, way deep down. The sad part is that it’s really hard to lose weight and a book like this just takes the wind right out of the sails of all those folks that were trying hard to fight the good fight and do the right thing. But everybody’s got to make a living, right? And if a guy would rather write about obesity than practice/teach law, that’s his prerogative, right? I mean, law is fricking boring! Who cares how many people get hurt by the nonsense that this guy has written?

Advertisements

About Dan Hunter

Retired software engineer wading through the obfuscation, confusion and contradiction that corporate and political funding of medical research along with ego over science has created.
This entry was posted in Diabetes, Diet, Nutrition and tagged , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s