FP4: Weird Science —

Fatty Power is our ongoing crowdsourcing* project.

In 2005, the CDC’s National Center for Health Statistics released a startling report: the Obese and Overweight BMI categories were associated with a net total of 26,000 annual deaths, which directly challenged the CDC’s revised report which claimed that these categories were responsible for 365,000 annual deaths.

Furthermore, this new report found that the Overweight BMI category had a negative mortality rate of 86,000 deaths annually, which gave birth to one of many “obesity paradoxes.” This paradox said that all-cause mortality was lower for those in the Overweight BMI category than the Normal category.

Lead author Dr. Katherine Flegal (who was among the first to raise the alarm on the initial increase in both adult and child obesity rates in the 1990s) presented the findings and was roundly attacked for health heresy. Her crime? Daring to suggest that the Normal BMI category wasn’t the ideal.

Flegal relied on the same data sample she had been using for decades, the National Health and Nutrition Examination Survey (NHANES) III, which used a series of mobile health units to track the health and nutrition status of adults and children beginning in 1971. These mobile health units weighed each participant, as well as tracked their health over the decades.

One month after publication of Flegal’s report, Dr. Walter Willett (and his amazing mustache), Chair of the Department of Nutrition at the Harvard School of Public Health, hosted a symposium to refute Flegal’s research. The panel claimed that the mortality rate Flegal’s team came up with did not take into account chronic disease in the lower weight categories, something known as reverse causation.

Willett vehemently refuted Flegal in the Journal of the American Medical Association, claiming “lean persons are a mix of smokers, healthy active persons, and those with chronic illness (due to the direct effects of disease on weight and sometimes purposeful weight loss motivated by diagnosis of a serious illness).”

Flegal and her team forcefully responded with evidence that they indeed accounted for reverse causation (and later went on to provide substantial corroborating research to bolster her work). Meanwhile, the CDC quietly accepted the new, lower mortality rates.

Willett’s attempt to discredit Flegal has mostly fallen flat and the obesity paradox continues to perplex researchers who adhere strictly to the “Fat is Always Bad” belief system.

Well, fast-forward five years: Willett’s back and he’s pissed.

Willet is one of 33 co-authors involved in a study released last Thursday which finds that (surprise surprise) being in the Overweight BMI category WILL KILL YOU NOW. Right now! So chop off your fucking arm and get your BMI into the Safe Zone!

And as an interesting side note, the following day, a panel of FDA advisors voted in favor of lowering the minimum required BMI for lap-band surgery:

About 15 million of them meet criteria for gastric banding surgery under existing guidelines, which say a person should have a body mass index of 40 or higher, or a BMI of 35 or higher if the person suffers from a weight-related medical problem such as diabetes or high blood pressure.

But here’s the kicker:

The FDA meeting was held at the request of Allergan, Inc., the manufacturer of the Lap-Band device. Allergan also conducted the study presented to the advisory committee.

…

Dr. Karen L. Woods chaired the panel meeting. Dr. Woods is a stockholder in Allergan, Inc. but was granted a waiver by the FDA. She did not cast a vote on the new guidelines. In a statement regarding her waiver the FDA said, “Dr. Woods was selected because she has expertise in therapeutic endoscopy, which is important for the deliberations of the panel, and has experience serving as a chair.”

Okay, so let me get this straight: Allergan requests that the FDA discuss whether to increase Allergan’s potential client pool (“If the proposed BMI guidelines are approved, Allergan says more than thirty million Americans would have BMIs within the newly eligible range.”); Allergan conducts the study the committee relies on; and an Allergan stockholder chairs the meeting, essentially controlling the discussion and direction. But because the chairwoman doesn’t vote, this is considered a fair hearing?

But I digress.

The FDA hearing is simply circumstantial evidence that all of these anti-obesity forces are colluding to drive us toward a single theory of obesity, which is not substantiated in the least. That theory is that fat in any form is a toxic substance and having any amount in excess will reduce your lifespan, not to mention your quality of life.

This theory refuses to take into account the abundant evidence, led by decades of research from Dr. Steven Blair, which suggests that lifestyle, rather than BMI, is the determining factor in overall health and mortality. So, a morbidly obese person who exercises will live longer than a thin person who is sedentary.

Willett’s theory refuses to take these factors into account, since this theory almost always comes attached to Calories In/Calories Out (CI/CO): if you’re morbidly obese and you exercise enough to be healthy, then you wouldn’t be morbidly obese.

That is why I believe that the Willett vs. Flegal fight is a proxy battle for CI/CO. On the one hand you’ve got Willett and his team making these broad claims that being thin or “Normal” is the best way to live a long and healthy life (i.e., thin people are more likely to eat healthy and exercise since they have a negative energy balance (they consume less and expend more calories), which keeps them thin), while Flegal and her team have found that BMI, or weight, has little to do with health (i.e., anyone who exercises and eats healthy foods can be healthy regardless of how these lifestyle choices impact their weight).

But one thing that Willett’s study did not control for, as Pattie Thomas (aka Fatty Pattie) explained to me:

People who diet and have yo-yo weight gains/losses have poorer health outcomes. This has been shown in other studies. One of the reasons that you don’t see studies controlling for this is that a group of people who have higher BMIs who have never diet would be very, very small. Dieting is so pervasive in our culture that it would be almost impossible to do a large-scale review study.

She also raises issues with the fact that Willett’s study is meta-analysis (as opposed to Flegal’s mobile health units), which means that his team analyzed large sets of research to look for patterns. She explains:

Pooling long-term studies and imposing a not-originally-planned for analysis can bring up a lot of problems. For example, say one of the 19 studies asked people if they smoked where another study got more into depth regarding second-hand smoke and found that a number of people were exposed to smoke even though they didn’t smoke. So who got excluded in this study? Only people who didn’t smoke? What if you wanted to exclude people who were non-smokers but were exposed to smoke in their jobs or their households. Well, you couldn’t do that with just one study without trying to do so with all the studies. Different longitudinal studies ask different questions and have different definitions of health, illness, exposure to risk factors, etc. So without seeing the full data report, the 13% becomes questionable. There are slew of alternative explanations to their assertion that the risk is higher.

Thanks to Michelle Allison (aka The Fat Nutritionist), I now have a full copy of the original report and I read through it last night.

Fair warning: I am not science-y in the least, but I do have common sense. So what I noticed may be a red herring without verification from our more technical readers. That being said, I found myself questioning quite a bit in the study.

First, a question. What is the difference, statistically speaking, between “controlling” and “adjusting”? This study says that they adjusted for “age, study, physical activity, alcohol consumptoin, education and marital status.” Is that different than controlling for them?

Second, in the “Inclusion Criteria” section, they write:

Key variables (height, weight, smoking status, and preexisting conditions) not available at baseline were later collected on questionnaires, so the baseline was redefined as the later date.

So, much is made of the baseline weights, but the baseline seems to shift based on available data. Would that impact results?

An interesting thing about this report is that it goes to great pains to reduce concerns about the increased mortality they found for BMIs less than 20. The hazard ratio (the risk of mortality) for Underweight (1.45) is comparable to the hazard ratio for the upper end of Overweight (1.44) between the ages of 20-49. Even so, the authors all but diagnose the Underweight category has suffering from some mysterious illness that couldn’t possibly explain those in the Overweight category.

They even go so far as to discuss the mortality difference in active and inactive Underweight people:

The hazard ratios for all BMI categories below 20.0 were also lower for participants who reported higher levels of physical activity than for those who reported loser levels, although the trend was no significant (P=0.14). The hazard ratio for a BMI of 15.0 to 18.4, as compared with 22.5 to 24.9, was 1.62 (95% CI, 1.38 to 1.91) versus 1.22 (95% CI, 1.02 to 1.46) for those reporting low versus high levels of activity.

I don’t have the Supplemental Appendix, so I’m not sure how this impacted the Overweight or Obese categories, but the fact that they include this information and zero information on the impact of activity levels in fat people raises my suspicions.

The study also takes swipes at Flegal’s study, taking great pains to point out that Flegal’s research included only a fraction of their data (because in statistics, more is always better, right?):

In contrast, analysis of NHANES data and the Canadian National Health Survey, which included smokers and persons with preexisting diseases [ed. note: except they didn't, as you recall] showed that being overweight was not associated with an increased all-cause mortality. These studies were smaller than our pooled study, with only about 11,000 deaths combined (7% of the total deaths in our study), so it is unlikely that their inclusion would have altered the main results of the current analysis.

Shorter Willett: Our study wins because we have more dead people!

But the thing that, to me, seems to indicate that this study is a farce, a charade, and a snipe hunt all wrapped up into one, is saved for the very end of the report. See if you can pick it up:

The principal limitation of our study is its reliance on height, weight, and preexisting conditions at a single point in time. As explained above, changes in these factors may contribute to the change in hazard ratios over time, but without repeated measures of these factors, we cannot assess their relative contributions.

Emphasis most certainly mine.

So, what this study really says is that if you, at some point in your life, are Overweight or Obese, then you are more likely to die than if at that exact same point in your life you are Normal.

In other words, this study is based on the weight of people from a single day, as opposed to examining the pattern of weight over time, and has determined that if you were ever Overweight or Obese, then you have a higher risk of dying.

That, to me, makes this report not only useless, but completely and utterly laughable. And this is supposed to be a response to Katherine Flegal’s study that actually weighed its participants, actually tracked their weights and health statuses over time?

This is Walter Willett’s big “Told You So” to the fatty heretics? A study that relies on self-reported weights, heights and diseases? A study that excludes non-Hispanic whites? A study that takes great pains to minimize the health risks of the Underweight? A study that set out looking for a correlation between Overweight and mortality, and found it thanks to a team of 33 researchers?

This isn’t an epidemiological survey, this is a vendetta, pure and simple.

And with that in mind, I want to take the dissection a step further.

Willett’s conclusions are based on 19 previous studies. I want to dissect those studies to see what they are made of. I want to invite our scientific readers to help us find out just what kind of foundation these results are based on.

This study is huge. It will be cited for years to come as evidence that ALL FATTIES must lose weight (preferably via lap band surgery). So, we must do something today to make sure that anyone who cites this study will also have to explain the limitations, the flaws, the holes and the contradictions.

I have created a Facebook group to organize this effort. I will serve only as a coordinator, since my contributions are limited by my left brain. Hopefully, by crowdsourcing this effort, we can turnaround a response quickly and add it to the record of dissent.

The CDC has manipulated data to further its agenda in the past and this is yet another self-serving example of science and government bowing to the Diet Industry. You can help take down this Goliath, so please join us.

*Wikipedia defines crowdsourcing as “the act of outsourcing tasks, traditionally performed by an employee or contractor, to an undefined, large group of people or community (a crowd), through an open call.”