So, Will Obesity Kill our Children at a Younger Age that Us?

In Weight-Loss Maintenance on March 5, 2011 at 10:06 am

Today I’m going to blow more than 1,000 words examining a single sentence in the Linda Bacon/ Lucy Aphramor treatise, Weight Science:  Evaluating the Evidence for a Paradigm Shift.  This sentence appears in the final paragraph of the subsection entitled, “Assumption:  Adiposity poses significant morbidity risk.”   Before I present the sentence, some pre-amble (digression).

In this subsection Bacon and Aphramor cast doubt on the assumption that our fat is killing us.  They separate “causation” from “association,” and also talk about and footnote (to my satisfaction) the “obesity paradox,” a convincing pattern that has emerged in epidemiology that links “overweight” and “obesity” to improved outcomes and longer survival periods in the case of many critical medical battles. 

Why do we call this an obesity paradox?  Why does it surprise us?  I’d call in the “obesity survival rate no brainer.”  Isn’t it fully logical that a person could benefit from some added fuel if he or she is contending with a serious chronic disease, recovery from a violent medical event (such as a heart attack) or a radical surgery that involves a rehabilitation period in which appetite is suppressed?  If there is a medically induced period where the body must waste some of its stores, who’s gonna fare better:  the presumably unhealthy “obese” person or the “healthy weight” person with the 19 BMI?  I know where I’m placing my bets in Vegas. 

Nevertheless, despite the recent acknowledgement of this “obesity paradox,” and despite the US National Institutes of Health telling us we are living longer and more robustly now than in any preceding generation, we keep hearing that our fat is killing us.  Moreover, there is a recent insidious catch phrase that has emerged:  “this is the first generation of children that may have a shorter life expectancy than their parents.”  Huh?  As Bacon and Aphramor point out, this statement continues to appear in popular mainstream media and with the imprimatur of a former Surgeon General.  But, is it true?  It’s alarming to think our children will die younger than us!

According to our authors this notion originated with a journal article by Olshansky et. al. A Potential Decline in Life Expectancy in the United States in the 21st Century.  Specifically, (drumroll, please, for the Bacon/Aphramor sentence du jour):

“When citation is provided {for impending decreased life expectancy}, it refers to an opinion paper published in the New England Journal of Medicine, which offered no statistical evidence to support the claim.”


Is that a fair characterization?  Did Olshansky et. al. really offer NO statistical evidence?  This off-handed statement deserves some examination.  I apologize, again, for only being able to get you through to the Abstract.  A source (thank you to You Know Who You Are) has provided me the full manuscript.  

As I analyzed Olshansky, et al., I think it’s an overstatement that the article relies on NO statistical evidence.  In its six pages of narrative and among its 67 sources are some statistical references.  It does, however, rely on old and disproved statistical evidence, such as studies by Allison et. al. and Mokdad, et. al. that each estimated annual excess deaths from obesity at more than 300,000.  These studies were convincingly refuted in April of 2005 (one month after the Olshansky article was published in NEJM) by Flegal et. al. in JAMA.  The Flegal study put the number of excess deaths at closer to 26,000 per year.  All of these studies, however, were only showing association and NONE analyzed whether excess deaths were really caused by obesity. 

I tend to think that if there are excess deaths associated with obesity, then most are not caused by the obesity itself, but from co-morbidities (as Bacon and Aphramor point out) or, worse, may be attributed to the dangerous countermeasures — beyond even fad dieting and yo-yo weight cycling — that fat people employ to reverse their obesity (something that Bacon and Aphramor should have pointed out).  By way of methodology, the three studies on excess deaths – Allison, Mokdad and Flegal – noted weight data collected on people in a particular time frame, then reported by BMI category which subjects had died ten years later (from all causes).  Allison and Mokdad discounted deaths from smoking, which resulted in suppressing the number of deaths of the trim people (who may have been using smoking as an appetite suppressant and/or were experiencing the body-wasting effects of smoking-related cancers).  Excess deaths among fat people, then, were artificially inflated by comparison.  Simply, it was bad science, and Flegal corrected it in order to assure we compared apples to apples (better science). 

But was Flegal’s study even a fair assessment of excess deaths from obesity, or might it have been showing us something else?  If we’re trying to understand whether obesity is fatal, I submit that rather than tweaking the statistics to suppress the death rate among the nonsubject group (as Mokdad and Allison did), we should look at whether there might be causes, other than fat itself, that might be inflating the number of deaths among the subject group.

During the subject study period, there were many outside elements that were likely killing fat people, and not trim people. The weight-loss drugs Fen-Phen and Redux enjoyed their hey day and were pulled from the market during that time, for example. Moreover, this was a time when weight-loss surgery was younger and even more dangerous than it is now. While all the studies (even Flegal) found excess deaths among obese people* ten years from the date they were weighed, shouldn’t we ask how many of those 26,000 excess deaths might well be subtracted since they were the result of interventions that artificially inflated the number of deaths among fat people?  In addition to Fen-Phen and Redux, what about over-the-counter preparations, like Metabolife?  Who would have taken those and how might they have been affected?   

I appreciate that Bacon and Aphramor identified the Olshansky paper as the source of the mythology that our children are going to die younger than we are.  Every time I’ve seen that strange quote I have wondered where the heck it came from, since it just doesn’t square with other studies on longevity.  Bacon and Aphramor are also right to indicate that the Olshansky paper is weak, and they would have been well within their rights to reinterpret the data it’s based on.  However, to say it offered NO statistical evidence was both arrogant and ironic, since their own essay is also an “opinion paper” (manifesto?) perched on limited evidence-based support.  As Living400lbs points out in the comments on our last post, “The papers they cite as specifically HAES-focused were only in obese women, sometimes with an eating disorder history. No studies focused on HAES in men.  No studies focused on people with diabetes, PCOS, hypothyroid or a history of heart disease.”

Whether obesity itself kills us is still up in the air.  I suspect that some forms of obesity are life-threatening and may merit treatment beyond a HAES model.  On the other hand, a vast number of people could benefit from a HAES approach to their health.

Sadly, we aren’t at a place in our discourse where we can parse close enough to find out which obesities should be treated and which should be managed with HAES.  Dr. Yoni Freedhoff is right that obesity research is in its infancy (see comments).  At this time, many scientists have embraced their hypotheses with a zeal that has morphed into bias, and these scientists just can’t write objectively.  (I wonder whether it is fair of us to expect them to, given how nascent this area of research still is.)  Unless they have had a change of heart since their paper in 2005, Olshansky, et. al., as well as Mokdad and Allison, are clearly what I call “fit in” scientists and Bacon and Aphramor conform to my nerd model.  They are advancing the debate, but awkwardly, often angrily.  Lone Wolf scientists like Flegal et. al., acknowledge that we are still crawling and it will be a while before we creep, then walk.  They implore us simply to compare apples to apples, for the time being, in broad strokes. 

I look forward to a day when we are sprinting about the orchard comparing Braeburns to Fujis to Jonathans to Galas, and not judging any of them to be lazy, ignorant, unmotivated or living the wrong “lifestyle.”

*Flegal did not find excess deaths among “overweight” people, by the way, and identified that the “ideal” BMI for increasing life expectancy at 10 years was around 27.

  1. I am totally with you that there may be other causes aside from fat itself that associates obesity with morbidity. One you didn’t mention that I think is critical is the effect that obesity stigma or discrimination has on health outcomes (see

    That said, I find the “first generation of children” statement entirely plausible. But not from obesity … I believe that the same thing that primarily causes obesity (our industrial diet) is the real culprit, and if anything, the prevalence of processed foods in our lives will in fact lead to higher rates and earlier diagnoses for diabetes, heart disease, cancer etc. Not a pretty picture considering our looming problems in health care cost and availability!

    One other thing that I also think bears lots more investigation is the applicability of Pottenger’s Cats research. I know that some people think his research was flawed, but I do think there’s something compelling in the idea that maternal health (particularly during pregnancy) can predispose a person to being more prone to these lifestyle diseases. So the generation of mothers who were raised on industrial foods could conceivably be setting up their kids to be disadvantaged compared to a generation ago.

    Re which obesities should be treated, I’m certainly encouraged by the fact that there’s so much research going on in terms of inflammation and the role of abdominal/visceral fat. The only scary part is that our only model right now is intervention by pharma. We could go thru many more Phen-Fens (with similar consequences) before really connecting the dots.

  2. Beth, have I told you recently that I think you’re brilliant? Amen on all counts.

  3. Wow.
    Just wow.

    Thank you for looking at that one line in such depth.
    From a public health perspective (I’m an MPH), I tend to think that poor health, obesity, and the social determinants of health are conflated. When we have a food industry that engages in what I consider predatory practices toward people with less education and less income, no regulation on food advertising, limited education about nutrition, plus economic, racial, social disparities that lead to health disparities, plus policies that protect the wealthy and punish the poor, we end up where we are now. There are other influences in the picture. But to then try to pull out only fatness as the primary factor contributing to ill health and early death — that’s so reductionist as to be nearly meaningless.
    I support a health- and patient-centered approach, which sometimes may mean attention to weight when its called for, but might also have to do with looking beyond the individual to see what the barriers are to focusing on improving health. Even HAES doesn’t address the barriers to physical activity and nutrient-filled eating that people experience. Childcare for single parents or kinship caregivers so they have time to be physically active or shop for or prepare healthy food is one gigantic gaping hole in the individual responsibility for health paradigm.

    • To the question: Will obesity kill our children at a younger age than us? I would answer: “Why start with a foregone conclusion?”
      If the question is: “Will our children have shorter lifespans than our own?” Then if the answer is “yes,” the next question is “Why?” To jump to the answer being obesity is… I don’t know what to call it… skipping several key steps?

  4. Frankly, with the growing social and financial inequality in the US, I wouldn’t be surprised if the average life expectancy decreased – but I doubt it would be due to obesity. Obesity does make a handy scapegoat, though doesn’t it? The demographic groups that are less likely to have medical insurance are also more likely to be fat – and more likely to live stressful, financially insecure lives without the time for physical activity and or the time/money to “eat healthy.”

    All of those factors probably have a more pronounced effect on mortality and morbidity than BMI does.

  5. I’m in love with your essay and the comments.

    We are witnessing the inevitable result of cultural beliefs about knowledge itself, about the ways in which human beings create meaning together.

    The World Health Organization tracks deaths from malaria each year, close to a million human beings dying from the disease yearly (give or take a couple hundred thousand), with hundreds of times more than THAT suffering from the terrible symptoms (but surviving). Malaria mortality epidemeology is so much more clear cut and measureable than any strained attempts to link obesity with mortality. And yet. The media (and I include high falutin’ peer reviewed journals in that category) remain focused on *obesity deaths*. Wow. What kind of minds do that without hesitation?

    It is like some kind of bizarre mass codependency in which particular groups are constructed (“the obese”) and made to seem *real* (reified); then each group is framed as another mass of units that need to be, and supposedly CAN be, controlled (or “managed”). The researchers, writers, and all who do not stop to question the underlying assumptions, etc, believe *we* must attempt to control the behavior of THIS group of people (such as people with supposedly larger body mass than is deemed acceptable), like some giant game of billiards using the laws of physics(!!!) because they (we) feel powerless…over so very much.

    It is a kind of projection of internalized powerlessness and fear, onto the *external* world.

    • DeeLeigh, RNegade, you make me feel “called” to protest somehow. This blog helps me release steam, but it does seem like a small gesture in the grand scheme of things. Indeed, obesity is the scapegoat for HUGE social and medical inequities.

  6. In times of social unrest and extreme inequality, we need scapegoats. The Turks had their Armenians, the Germans had their Jews, the Hutus had their Tutsis. It’s harder now to find a scapegoat without being accused of some kind of “ism”.

    I don’t want to compare to the overweight to any of the above groups, but think about it. The overweight can be hated and despised for their supposed weakness and even blamed for how they are supposedly overburdening the health care system. Add to that the fact that (as others have pointed out), overweight and obesity are often associated with low income and education and you get extra bang for your buck. You can despise the poor (who also have only themselves to blame)and the overweight in one fell swoop.

  7. “this is the first generation of children that may have a shorter life expectancy than their parents.”

    I don’t put much stock in that statement, if only because it’s been floating around for a few decades. It’s meant to frighten us, because we’re clearly not panicking enough about this horrible obesity epidemic.
    When you think about it, historically it’s been the norm for children to have shorter life expectancies than their parents. That’s why people tried to have so many – in hopes that some would survive to adulthood.

  8. I was just wondering what you thought of this:

    “How the war on obesity went pear-shaped

    “With increasing evidence that a big belly probably won’t lead to an early death, it’s time to call a ceasefire.”

    I agree with you that there’s almost certainly more than one kind of obesity, with different levels of health risks, but I’ve never liked the “apple versus pear” choice that many people make. One of my friends is “death fat”, with a huuuuuge belly, and is currently trekking India for charity. My obesity, on the other hand, is definitely causing health problems for me in and of itself.

    • Welcome to the blog, Liz. I don’t know what to make of that study. It seems that everyone is jumping on it to prove a point that it may not have been intended to make, and that’s dangerous.

      I think the study was just saying that body shape doesn’t add any relevant information, if you already have good information on a patient, such as age, sex, smoking status, baseline systolic bp, history of diabetes, and total and HDL cholesterol numbers. That said, the apple body shape may also be associated with a lot of those relevant factors. So, the take-home message here: if I were apple shaped, I’d want to ask my doctor to check for the relevant factors routinely, just to assure me that all was well. (Actually, all people should have those factors checked.) However, I would also be disgusted if the doctor assumed that I was not okay, judged me, simply because of my body shape, and recommended treatment without checking the relevant factors.

  9. I believe you missed the point of our article in 2005, and once again here. First of all, it is not possible to provide “statistical evidence” for a future event involving health and longevity — this is a ruse. In our NEJM paper, we set out to estimate the negative effect of obesity on life expectancy at that time — our conclusions were correct then and they are still correct today. They were not based on disproved statistical evidence as you suggest, nor did the Flegal article disprove anything in our article. You are inappropriately mixing an estimation of the current negative effect of adult onset obesity and its impact on life expectancy (which we documented in our NEJM paper), with the possible future effect of the childhood obesity epidemic on future levels of life expectancy (which was speculative). The negative longevity effects of childhood obesity take time to be observed — this is referred to as a latent effect. What you have suggested is equivalent to saying that smoking appears to not be harmful to longevity among a group of 20 and 30 year old people who started smoking before the age of 10. Smoking does kill early, and quite dramatically, it’s just that you won’t see this until these smokers begin to reach the ages at which this risk factor expresses itself. The reason we do not yet see the negative effects of childhood obesity on life expectancy is because not enough time has passed for the effect to be observed — this is especially true among younger children who will not be like the people who acquired obesity as adults — it will be much worse for these children because they will carry related risk factors for decades longer. In fact, we are drafting a new manuscript now that documents the evidence that has emerged since our 2005 NEJM paper indicating the negative effects of obesity acquired in childhood are already being observed among people now in their 20s and 30s.

    We explain the problem of the “latest effect” in our latest article just published yesterday:

    I hope this helps clarify this issue.
    S. Jay Olshansky, Ph.D.

    • Quick reply for the time. Am on my way out the door. First of all, thank you for taking me seriously enough to respond. Will reread the 2005 piece and your new piece and will respond. It may take a while. I’m still digesting a couple of pieces for Dr. Katarina Borer, using my lay person eyes and maintainer sensibilities. I promised her I would.

      With regard to your work, don’t get too enthusiastic. I’m pretty cynical. But I also try to be fair. As you see, I’m not 100% on board with the HAES crowd either, and haven’t cut Bacon and Aphramor much of a break. Manifestos of any variety, this early in the research (and it is nascent!) rub me uncomfortably.

      Don’t have the link handy, but in my October Archives is the lens with which I analyze obesity experts — it’s the Revolutionary, Breakthrough method for Evaluating Experts, in fact. Those two adjectives are in its title at any rate. It’s a bit tongue in cheek, but it puts my biases on the table.

      Thank you for visiting. And thanks for the link. At present it’s even letting me through to the full manuscript. Hooray!

Leave a Reply

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

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

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s

%d bloggers like this: