It has been interesting to see the disproportionate flow of page views to Basic Weight and Obesity Facts from Tumblr and Reddit. The deconstruction and criticism has been enlightening and helpful for me, but it also points to the fundamental human flaw that we all have: if we believe something, then we will shape the facts to better support our belief.
The individuals who are critical, dismissive and angry at the information presented in Basic Weight and Obesity Facts believe that I am misleading individuals. For them, obesity is not just a disease, but it also causes disease and the cure is weight loss and exercise.
There was a time when I would disinterestedly agree with them. Let’s be honest, obesity has no bearing on my life and so I was as lazy as the next person and simply assumed that people are obese because they eat the wrong foods, eat too much of those wrong foods, and fail to get out and move around enough too.
The advantage I have over everyone who has a vested interest in believing that obesity causes illness is that I have no vested interest. As I started researching the hard data I discovered that my mainstream and unchallenged assumptions were dead wrong.
Not only is obesity not a disease; it does not cause disease. In fact, it is very likely that becoming fat (i.e. the fat organ increases in size) is the body’s protective response to the appearance of chronic disease for completely distinct and far more complex reasons than burgers and couches.
And those distinct reasons would likely require much more investigation and discussion about increasing social inequality, prevalence of psychoactive drugs (including antidepressants), and the presence of some 77,000 untested chemicals in everyday household and personal items.
Yet another public ‘education’ campaign to weigh children in schools or seriously entertain the idea that parents should be had up on charges of child abuse because their children are obese, keep us searching for keys we dropped in a dark alley by groveling around a lamppost that’s actually around the corner and down the street from that alleyway [Streetlight Effect].
To address the metabolic conditions in our populations that cause the fat organ to have to increase in size would negatively impact the juggernaut of transnational shareholder return on investment.
I see trends continuing much as they have been when it comes to the so-called war on obesity. Our culture finds it more expedient to blame the victim rather than risk further job losses or commodity price increases— the two twin specters of doom that industrial lobbyists will raise as threats whenever there are even tentative efforts to find true causation.
But let’s get onto some further facts on obesity...
Mainstream Press Bias
While many on Reddit and Tumblr took care to joust with peer-reviewed published materials, most people will rely upon popular media coverage of peer-reviewed published material to frame their understanding of a condition.
“Media outlets must make choices when deciding which studies deserve public attention. We sought to examine if there exists a systematic bias favoring certain study design in the choice of articles covered in the press. Our results suggest such a bias; the media is more likely to cover observational studies and less likely to report (randomized clinical trials) RCTs than a reference of contemporary articles that appear in high impact journals. When the media does cover observational studies, it selects those with lower sample sizes than observational studies appearing in high impact journals.
While it may not be surprising that the media tends to select articles outside of the highest impact journals, in doing so, they preferentially choose articles lower in the hierarchy of research design, thus favoring studies of lesser scientific credibility. If anything, as top newspapers have their pick of all original articles, not just those selected by high impact general medical journals, newspapers could choose to cover the most credible studies, i.e. large, well-done RCTs. Instead, collectively they appear to make an alternative decision.”
[S Selvaraj et al., 2014]
It is thanks to the Health News Review blog that I was made aware of the above paper and also that the journal news releases themselves may be a big part of the problem: misleading BMJ news releases may be one reason journalists report on more observational studies.
If a friend forwards you an article they’ve found that confirms obesity is a killer, then go to the source of whatever published research is being referenced in the article. That’s not always easy to do, as these articles will quote the researchers, but fail to provide the title of the paper or where the paper has been published. If you don’t want to be bothered doing what I am often forced to do, which is take the time to search based on what I can glean from the article, then just dismiss it out of hand if the journalist has failed to provide clear references.
By the way, for those unaware, references for blog posts on this site are found in the References section and you can search the entire index and be taken to the original paper there. That is why you will not find complete references at the end of each individual blog post here.
The References section still has some link errors that I need to clean up, and is not always completely up-to-date. Right now I am behind on updating them. If you find that something I mention in a particular blog post cannot be located via the References then either use the comments feature (if the blog post is new) or use the Contact form – it will be much appreciated!
So first up on the obesity bullsh*t trail is an example from a YE community member who received an article from a friend to which I responded in detail within the forums. I have decided to excise my response and expand on it. I will unpack the data, explain the necessity of reviewing actual numbers, and describe what epidemiological studies can and cannot do for us.
For those of you unfamiliar with the term “obesity paradox” refers to experimental data that suggest those with various chronic health conditions, such as heart disease or diabetes, are less likely to die or suffer more serious complications when they are obese and have these conditions than when they are average-weighted and have these conditions [JP Curtis et al., 2005; W Galal et al., 2009; A Oreopolous et al., 2008]. There is a lot of money being sunk into an effort to disprove this so-called obesity paradox.
And, no surprise, the above material got huge, heavy mainstream media rotation. Now bear with me as I unpack the data behind the rather wishful-thinking conclusion stated in the above headline.
So, this is an epidemiological study. All a study like that can do is suggest correlations, not causation.
Nothing wrong with epidemiology, but mainstream articles like the one above do not explain that the correlations revealed still do not uncover what is causing what.
There is a rather important and critical piece of information in this review of mortality outcomes for 11,000 odd male and female nurses diagnosed with Type II diabetes mellitus:
“[The study] looked at deaths according to how much people weighed when they were diagnosed with diabetes." (emphasis mine)
Fat is not a storage unit; fat is a major hormone-producing organ in the body [EE Kershaw, JS Flier, 2004; JN Fain et al., 2004; RS Ahima, JS Flier, 2000].
The study suggests that those who are heavier when they develop diabetes are more prone to die than those who are not heavier when they develop diabetes. These are correlations, meaning that when one thing occurs (diabetes and obesity) then another thing occurs (death) in such a way that chance can be ruled out as being the reason.
Is this correlation strong or weak? I’ll get to that in a moment.
But as far as whether obesity causes diabetes or obesity causes increased diabetic-related deaths, this study provides no such causative results.
It is possible that the subjects who were heavier at the time of diabetic diagnosis had fat organs that were already trying to respond to the metabolic development of a more severe case of diabetes. In other words, maybe they would have lived even less time had they not been as large as they were at the onset of diabetic diagnosis. We don’t know.
The mortality data are important in studies like these, and yet often they are not provided in mainstream media coverage at all.
Unpacking the Data
So here goes, we'll unpack the actual dreaded "J-shaped" curve that shows those who are underweight or "obese" have greater mortality risk than those who are average or "overweight" by actually going to the study overview itself.
There were 11,427 patients assigned to this study with Type II diabetes mellitus.
In the 15-year study period there were 3,083 deaths. If you were any of those people in that study, you had a roughly 3 in 4 chance of still being alive at the end of the study. Yay modern medicine.
Alright, now we'll unpack the BMI to death hazard ratios.
But before I even get to that, I should point out that I do not know if the 3,083 deaths were directly attributable to complications from diabetes or not. But for the sake of trying to paint the worst picture we can, let’s assume everyone actually did die of complications from diabetes.
BMI 18.5-22.4 (at the time of diagnosis) HMR 1.12
BMI 22.5-24.9 (at the time of diagnosis) HMR 1.00
BMI 25.0 to 27.4 (at the time of diagnosis) HMR 1.16
BMI 27.5 to 29.9 (at the time of diagnosis) HMR 1.21
BMI 30.0 to 34.9 (at the time of diagnosis) HMR 1.36
BMI ≥35.0 (at the time of diagnosis) HMR 1.56
What’s HMR? Hazard mortality ratio.
It is called a hazard mortality ratio rather than a standard mortality ratio because this review has internal comparators and the data were not compared to a population wide standard mortality ratio (SMR). But the concepts are somewhat interchangeable and using HMR instead of SMR may avoid death rate underestimations [TR Card et al., 2006].
Here's how it works: the researchers determined that however many people died in the study period that were BMI 22.5-24.9 were dying at an expected rate. That translates to an HMR of 1.00.
Let's just say for argument's sake (as I don't have access to the full study with actual subject breakdown) that there were 1,904 people who were BMI 22.5-24.9 in the study and that 514 of those in that bracket died within the study period (about one quarter of the total). 1,390 in this bracket were alive at the end of the study.
61 more people die who began at BMI 18.5-22.4 than those at BMI 22.5-24.9.
1,329 people in this bracket were still alive at the end of the study.
82 more people die who began at BMI 25-27.4 than those at BMI 22.5-24.9.
1,308 people in this bracket were still alive at the end of the study.
107 more people die who began at BMI 27.5-29.9 than those at BMI 22.5-24.9.
1,283 people in this bracket were still alive at the end of the study.
185 more people die who began at BMI 30-34.9 than those at BMI 22.5-24.9.
1,205 people in this bracket were still alive at the end of the study.
287 more people die who began above BMI 35 than those at BMI 22.5-24.9.
1,103 people in this bracket were still alive at the end of the study.
If you were part of that study you were more likely to be alive at the end of the study no matter your starting weight than not. Or, 7,618 out of the original 11,424 are still alive.
If we do assume equal numbers of individuals were in each BMI bracket, then the relative risk increase of mortality from the standard group of BMI 22.5-24.9 to those who were above BMI 35 is about 9%.
Another study showed that diabetics who were at greater than BMI 29 who attempted to lose weight managed to lower their relative risk of mortality to 0.84 [N Chaturvedi et al., 1995]
Yet another study confirms that intentional weight loss for those classified as unhealthy can lower mortality risk to 0.87 [M. Harrington et al., 2009].
The 'whopping' improvement seen with weight loss for these diabetic subjects was achieved with an intentional loss of as little as 1 kg or 2.2 lbs.
Keep in mind that standard mortality risk is always 1.00 so the difference from 1.00 to 0.87 or 0.84 is statistically relevant, but from a quality of life perspective may not really be worth turning down the chocolate cake.
And here's what we don't know from looking at these data:
Were patients who were at heavier weights when diagnosed with Type II diabetes dealing with more severe metabolic issues that would have resulted in poorer outcomes no matter the weight? We don't know.
Given that Type II has a variable prodrome* (often controversial, but referred to as metabolic syndrome) then were the patients who were heavier at the time of diagnosis getting heavier prior to diagnosis? And if so, was it because the fat organ was trying to maintain homeostatic balance and therefore did they actually live longer than the arbitrary "at diagnosis" marker would suggest precisely because their fat organ was already working to avoid the onset of full blown Type II diabetes? We don't know.
*prodrome: the period of time during which subclinical symptoms are developing or present prior to the onset of a full-blown clinical state.
It doesn't actually tell us that above-average weight is not health or life protective, only that there are HMR increases that result in 287 more individuals dying in a 15-year period who are above BMI 35 at the time of diabetes diagnosis when compared to those who are BMI 22.5-24.9.
And then there is the research funding: Deirdre K. Tobias and her colleagues' work, referenced above, was generously provided for by the American Diabetes Association as well as the National Institutes of Health.
Here is a list of the corporate sponsors for the American Diabetes Association:
Abbott, Amlin, Lilly, Merck, Novo Nordisk, Sanofi, BD Boehringer Ingelheim, LifeScan, Cary's Sugar Free Syrup, Colgate Total, Domino Sugar and Stevia Blend, MedPeds, NutriSystem, Walgreens, Walmart, Boar's HEad, BeeWell For Life, Catherines, Curél, Dannon, Dr. Comfort, Equal, Gold's Gym, Janssen, Shire, Splenda.
So when you have a friend e-mailing you ‘proof’ of how deadly obesity is, know that were you even diagnosed with Type II diabetes at a BMI of higher than 35, the chance of you dying of an eating disorder still blows the slightly increased risk of dying of complications associated with Type II diabetes right out of the water. I provide the SMR for eating disorders later on in this post.
And furthermore, obesity does not cause diabetes. Whether the fat endocrine organ is taxed directly by unknown environmental stressors, or other organs or systems in the body are stressed by environmental stressors causing the fat organ to respond by becoming larger to try to modulate the impacts of those stressors elsewhere in the body, is unknown. Not to mention we also have yet to integrate the fact that there will inevitably be genetic predispositions for these kinds of metabolic chronic conditions as well.
Framingham vs. Flegal
Next up I will excise a response I made in the comment section of Basic Weight and Obesity Facts. But first I will provide more information on the oft quoted multiple analyses of the Framingham Study when compared to the largely unknown and solid work of Katherine Flegal and her colleagues.
Framingham Heart Study: Epidemiological Cohort Study
Ah Framingham, let me count the ways that these data have been manhandled and tortured into whatever shape best suits the concern du jour.
The first round of the Framingham Heart Study was initiated in 1949 and ran to 1966. Subsequent rounds have included the Offspring cohort, the Generation Three cohort, the Offspring Spouse cohort, the Omni cohort as of 1994 and the second Omni cohort in 2003 comprising the grandchildren of the original 5,209 participants from Framingham Massachusetts.
Well over one thousand peer-reviewed published papers have relied on the data from the Framingham Heart Study.
One of the most powerful outcomes of any epidemiological study (Framingham is an epidemiological study by the way) is that it can uncover correlations that take decades to show up given our fairly long lifespans as humans. Framingham’s biggest claim to fame was its ability to show a very strong correlation between smoking and heart disease.
But do you know how we discovered that cigarette smoking causes heart disease? By actually seeing the rates of heart disease go down as smoking cessation went up in our populations. And quitting smoking for those with existing heart disease is also correlated with substantial risk reduction for all cause mortality as well [JA Critchley et al., 2003].
Correlations That Are Not Correlating
“Great,” you say, “so then the correlations we see with obesity and diabetes will translate into causation once all obese people lose weight and diabetes rates inevitably drop through the floor as a result, right?”
Nice try. Problematically the rates of diabetes among non-obese and obese individuals are rising despite the fact that obesity rates have remained flat in most developed nations for the past 10 years.
The estimated prevalence for diabetes worldwide in 2010 is 6.5% and expected to increase to 7.7% by 2030 [JE Shaw et al., 2010].
In the United States (as a developed nation) the prevalence of obesity from 1999-2008 remained stable at 33.8% with the rate for women flat throughout that time frame and the rate flattening for men from 2005 [KM Flegal et al., 2010].
And a similar trend was seen for female and male children over the 12-year period of 1999-2010 in the United States. The prevalence of obesity for females aged 2-19 sat at 15% whereas prevalence in males of that same age range increased to 18.6% by 2008 and then stayed flat from that point onwards [CL Ogden et al., 2012].
And just to compare apples to apples, from 1995-2010, the age-adjusted prevalence of diabetes in the United States increased 82.2% [Centers for Disease Control and Prevention,2012].
Know what that means? There doesn’t appear to be any correlation for obesity and diabetes when the former condition stays flat and the latter increases exponentially.
Now, to be fair here, some of that massive increase in prevalence of diabetes has a lot to do with ever lowering test result levels being assigned as confirmed diagnosis of the condition. But, even if we assumed 50% of all new cases of diabetes are just over diagnosis, there is still an increasing prevalence of diabetes and a stable incidence of obesity.
Also diabetes mellitus Type II was always known to be prevalent in older folk and we have aging populations in all developed nations, so naturally the prevalence will increase as the Boomer generation ages. But that is exactly why the above data are age-adjusted of course.
But I digress. Framingham…
GOYRS: Know Your Risk
Here is an interesting conclusion from a peer-reviewed published comparative analysis of the Framingham data from the earlier period of 1952 to 1974 to the later time period of 1975 to 1998:
“The proportion of CVD [cardiovascular disease] attributable to DM [diabetes mellitus] has increased over the past 50 years in Framingham. These findings emphasize the need for increased efforts to prevent DM and to aggressively treat and control CVD risk factors among those with DM.” [CS Fox et al., 2007].
That statement is a misrepresentation of the actual data from the Framingham Study. It takes a moment to locate the sleight-of-hand wording, but it rests with “proportion”. And even then, it is still misleading:
“In these studies, the hazard ratios for diabetes as predictor of incident CVD were similar across follow-up periods, whereas the age- and sex-adjusted CVD incidence rate decreased by 50% between the earlier period (1950-1966) and the later period (1977-1995) of follow-up. This report ended a long debate by confirming that the decline in CVD rates observed in the West over the last decades of the 20th century had similarly benefited people with and without diabetes. (emphasis mine)”[AP Kenge et al., 2010].
Due to the massive amount of data collected, Framingham material was synthesized to produce the Framingham Risk Score (FRS). The FRS plugs in a bunch of your numbers (blood pressure, HDL levels etc. etc.) and then spits out your presumed risk of coronary heart disease. There is some credible academic discussion that the FRS overestimates coronary heart disease risk.
Recent research suggests that assessing coronary artery calcium levels via computed tomography scan (CTCA) provides more accurate risk assessment for coronary heart disease [MJ Pletcher et al., 2004]. However presumably a greater body mass index can create significant noise thereby lowering the accuracy of a CTCA for those above BMI 26 [H Alkhadi et al., 2008].
Meta-analysis suggests that elevated c-reactive protein is a modest predictor of coronary heart disease [J Danesh et al., 2004]. Then there is yet further research investigating the prediction value of measuring several other inflammatory markers (interleukin 6, serum amyloid A, fibrinogen…) for identifying the risk of coronary heart disease.
And of course there should be this much focus, investigation and money sunk into understanding the risks of coronary heart disease because it is, after all, our number one killer right?
597,689 people died in the United States in 2011 from heart disease. Next up is cancer, but don’t forget a further 129,476 died from stroke too.
Two million people died in the United States in 2011 (all cause mortality). The population in that country in 2011 was 311,800,000. The death rate was 7.9 people out of 1,000, or 0.7% [Centers for Disease Control & Prevention Fast Facts].
There is absolutely one death per person, but it’s not looking as if we are all rushing to the finish line there.
The chance of not dying of anything at all in 2011 was staggeringly high in the United States. I think I’ll call that the Get Over Yourself Risk Score (GOYRS). And GOYRS is even higher in most other OECD countries when compared to the United States as well.
Now onto Flegal…
National Health & Nutritional Examination Survey (NHANES): Epidemiological Follow-up Study
I am more partial to the work of Katherine Flegal and her colleagues in the review of NHANES. And while their most recent 2013 publication, where they performed a systemic review of published data, still finds mortality increases for Class 2 and 3 obesity, full paper here, the limitations of the findings are clearly spelled out in the comment section.
And we need to remember that "significantly higher all-cause mortality" is being used as a statistical statement.
The actual data from this review assigned BMI 18.5-24.9 as a hazard mortality ratio (HMR) of 1.00. That means that however many individuals died in the time period in that BMI bracket are considered to be the number of deaths expected.
The HMR for those in the "overweight" classification was 0.94 (slightly less chance of dying than average weighted folk). Class 1 obesity (up to BMI 34.9) had the same slightly less chance of dying than average weighted folk (0.95). Class 2 obesity (up to BMI 39.9) and Class 3 obesity (beyond BMI 40) were combined to have a hazard mortality ratio of 1.29.
That is most certainly statistically significant, but in real terms it means 29% more deaths in Class 2 and 3 obesity levels than average weight over the period of time in question.
If 3 average-weighted people out of 1000 die in an 18 month study period and that is our HMR of 1.00, then a 29% increase means that 3.87 people die out of 1000 if they are all Class 2 and 3 obese. That result is statistically relevant, but not exactly something that requires a real-world freak out.
Or let's say it's 30 out of 1000 people in a 15-year period as our HMR of 1.00. Then it's going to be 38.7 Class 2 and 3 obese out of 1000 dying in that same time period.
By comparison C Laird Birmingham's study on the mortality rate for anorexia nervosa [CL Birmingham et al., 2005] was 10.5.
Let's translate: in that 20 year time period let's say 30 out of 1000 non-ED general population people die (SMR =1.00), then in that same time frame 315 out of 1000 ED patients will die (SMR =10.5).
Another review of eight studies that assessed SMR for eating disorders arrived at an average SMR of 6.4 [S Neilsen et al., 1998].
Often the SMR for bulimia nervosa will be listed as much lower than that of anorexia nervosa in clinical studies, however recent published material is taking a much closer look at suicide rates associated with eating disorders (which can often be missed and therefore misclassified). And suicide rates are much higher for those with bulimia than anorexia [DL Franko, PK Keel, 2006].
The offspring of Denial and Fear is the intractable and infertile Dogma. Dogma is a set of principles laid down as incontrovertibly true. Society’s views of obesity are dogmatic and scientific study is being subverted to support those views at present.
Science is ostensibly meant to be a practice of inquiry— the very antithesis of dogma. You’ve probably heard the old adage: power corrupts; absolute power corrupts absolutely. Well not much about scientific practice in today’s world is anything other than an absolutely corrupted expression of those who wield absolute power.
I have a very uneasy relationship with scientific practice these days. My faith in this practice of inquiry is sliding around on rather slippery ground. I am like someone who clings to a belief in an ultimate deity who is above reproach, despite the fact that her faith in organized religion has been shattered.
I mention above the funding bodies supporting Deirdre Tobias and her colleagues, but what about Katherine Flegal and her colleagues? Flegal is a senior research scientist for the National Center for Health Statistics, Centers for Disease Control and Prevention. She is therefore primarily funded by Centers for Disease Control and Prevention and the National Cancer Institute, and some of her colleagues (depending on the paper involved) have received funding support from the National Institute of Diabetes, Digestive and Kidney Disease.
I would love to say that such government funding, as is awarded to the CDC through the United States Congress, is independent from corporate interest, but that would be an oversimplification of how the complex interplay of government and corporate interests are intertwined in today’s world.
Therefore, we could likely say that Flegal’s work is less likely to provide results that are supportive of corporate interests than the work of Tobias (where her funding relies more directly upon interested corporations). That is not to say that Tobias’ work is necessarily manipulated or fraudulent in any way— just that the risk of it being so is higher.
And no scientific inquiry today is completely independent from where the money originates. Money doesn’t talk; money is decidedly mute and yet phenomenally persuasive.
I may not be entirely comfortable depending upon peer-reviewed published papers to reflect unbiased scientific inquiry, but I am entirely comfortable stating that were diet and exercise the key to health and longevity, then we would have seen at least strong correlations by now (perhaps even suggested causation), and we have not— neither in scientific journals nor in real life.
Back In the Dark Alley
I refuse to grovel any more under the lamppost looking for keys that were clearly dropped around the corner and down the dark alley.
Do I know where the keys are? No, but maybe if I stand here long enough hollering down the alley for any passing scientist to start looking in the right place, we might finally see some progress in our understanding of numerous metabolic, yet largely survivable, conditions that plague many in our society today.
Dieting has a 95% failure rate. Persisting with the prescription of weight loss for ‘obesity’, or its supposedly tenuously correlated and completely uncorrelated chronic conditions, is like prescribing decapitation as a treatment for headache.
How are we able to collectively sleep at night when we actively blame the owners of a larger-than-average fat organ and discriminate against them as well?
And when did we lose all perspective such that we wring our hands over ridiculously overblown mortality rates for numerous chronic conditions as though they were equivalent to what real pathogens can do to us as a species?
Smallpox had a mortality rate of somewhere between 30-35%. And the worst mortality rate you can find by country for diabetes is Mauritius at 1.7% [World Health Organization, 2011].
There is one chronic condition that has a mortality rate pretty much on par with smallpox and that’s restrictive eating disorders— somewhere between 20 to 25%.
Awareness and Perspective
I have promised everyone a thorough blog post on orthorexia and all the restrictive diets du jour (paleo, raw, vegan…blah, blah) and although it is going very slowly, it is underway.
I know it is important to keep hollering in the alleyway, and that those attempting recovery from an eating disorder are unendingly assailed by all those groveling around the lamppost sure that the piece of garbage they have found is actually the keys that everyone is looking to find.
But for right now, as we enter February soon (Eating Disorder Awareness Week), please tell anyone who is not struggling with an eating disorder to stop misleading themselves and others about ‘healthy’ diets and exercise and tell them to calculate their Get Over Yourself Risk Score instead.
And calculating a GOYRS is easy to do: do you live in an impoverished developing nation, have smallpox, or have an eating disorder? If your answer is no, then stop shaming yourself and others with misconceptions of so-called healthy living and just go live your extremely favorable odds of a long and well-lived life.
I don’t have smallpox, a restrictive eating disorder and I live in a wealthy and developed nation, so I have the luxury of knowing I have a fabulous GOYRS.
If you do have an eating disorder, or suspect you do, then you have a genuinely deadly chronic condition and not the fauxpocalypse-inflated mortality rates associated with most other chronic conditions present in our society today.
I am not saying the chronic conditions in question are not extremely difficult to live with, or that they are not responsible for much suffering and reduction in quality of life, just that some overall perspective is required here.
As such, you don’t have to educate others on all things eating disorder (unless you want to), but you will have to accept a level of awareness and perspective that many others have the luxury of denying outright.