[Paleopsych] Skeptical Inquirer: Obesity: Epidemic or Myth?

[Todd I. Stark]

Some thoughts ...

In practical terms, it comes down to confusing two problems:  the 
problem of trying to apply statistics to individuals, and the problem of 
trying to inform public health policy with statistics.  As with 
environmentalism, the statistics on risks and incidence of obesity have 
often been interpreted in such a way as to make them seem worse than 
they are, usually with the well-intentioned goal of extraploating from 
real data and heading off worse problems.  However in the process, 
important distinctions often get neglected.

Two points in particular:

(1) some people have serious health risks related to obesity such as 
diabetes, sleep apnea, and heart disease, while others can carry a 
remarkable amount of body fat without much real health impact other than 
slowing them down a bit.  The serious health impact of body fat depends 
heavily on other risk factors.  This means that some non-trivial number 
of people will be unfairly treated as health risks when we create public 
policy that equates obesity with poor health, while for many it will be 
a legitimate connection.

(2) the health risks of overweight yield a completely different curve 
than those of morbid obesity, there is a sharp discontinuity.  Many 
people actually have fewer health problems when carrying slightly more 
body fat, at least until it becomes gross obesity.  There isn't really 
much health justification for representing mild overweight and gross 
obesity as points on the same continuum for purposes of risk factor 
analysis.  Some of the worst risk factors are not aggravated by body fat 
until it becomes fairly extreme.  It is simply not fair to judge a 
person's health by the fact that they are not lean.

Obesity is a legitimate health problem and contributes heavily to 
serious complications in many people and it is increasing in the U.S.. 
That much is not myth.  The measures we take to try to stem it are 
likely to unfairly affect people that are not unhealthy at their body 
weight.  That much is also true.  The tension of trying to interpret 
obesity as an epidemic or an overreaction depending on your goals and 
how you are affected seems hard to avoid.

kind regards,

Todd

Premise Checker wrote on 11/21/2005, 1:42 PM:

 > Obesity: Epidemic or Myth?
 > http://www.csicop.org/si/2005-09/obesity.html
 >
 >     New evidence shows that the obesity epidemic is not as bad as we have
 >     been led to believe. However, that doesnt mean that we should dismiss
 >     the problem either.
 >
 >     PATRICK JOHNSON
 >       _________________________________________________________________
 >
 >     You have probably heard that we are in the midst of an obesity
 >     epidemic. The Centers for Disease Control and Prevention (CDC) have
 >     been fervently warning that we are in imminent danger from our
 >     expanding waistlines since the beginning of this decade. However,
 >     evidence has recently emerged indicating that the CDCs warnings were
 >     based on questionable data that resulted in exaggerated risks.
 >
 >     This new evidence has led to a hostile backlash of sorts against the
 >     CDC. The editors of the Baltimore Sun recently called the earlier
 >     estimates the Chicken Little Scare of 2004. The Center for Consumer
 >     Freedom, a group that has long been critical of the CDC, declared
 >     unequivocally on its Web site and in print ads in several newspapers
 >     around the country that the obesity scare was a myth (figure 1). Even
 >     Jay Leno poked fun at the CDC in one of his Tonight Show monologues,
 >     making the observation that not only are we fat. . . . We cant do
 > math
 >     either. Not everybody believes the new data, however. Cable talk show
 >     host Bill Maher commented during an episode of his show Real Time
 > with
 >     Bill Maher about it being a shame that lobbyists were able to
 >     manipulate the CDC into reducing the estimated risk.
 >
 >     So which is it? Are we in imminent danger, or is the whole concept a
 >     myth? Looking at the scientific evidence it is clear that the extreme
 >     views on either side of the argument are incorrect. There is no doubt
 >     that many of our concerns about obesity are alarmist and exaggerated,
 >     but it is also apparent that there is a real health risk associated
 >     with it.
 >
 > The Controversy
 >
 >     Between 1976 and 1991 the prevalence of overweight and obesity in the
 >     United States increased by about 31 percent (Heini and Weinsier
 > 1997),
 >     then between 1994 and 2000 it increased by another 24 percent (Flegal
 >     et al. 2002). This trend, according to a 2004 analysis, shows little
 >     sign of slowing down (Hedley et al. 2004). The fact that more of us
 >     are getting fatter all the time raises a significant public health
 >     concern. The Centers for Disease Control and Prevention (CDC) began
 >     calling the problem an epidemic in the beginning of this decade as
 > the
 >     result of research that estimated 280,000 annual deaths as a
 >     consequence of obesity (Allison et al. 1999). Since then there has
 >     been a strong media campaign devoted to convincing Americans to lose
 >     weight. In 2003, Dr. Julie Gerberding, the director of the CDC,
 > made a
 >     speech claiming that the health impact of obesity would be worse than
 >     the influenza epidemic of the early twentieth century or the black
 >     plague of the Middle Ages. In 2004 the campaign reached a fever pitch
 >     when a report was released that increased the estimate of
 >     obesity-related deaths to 400,000 (Mokdad et al. 2004). Finally, in
 >     March of this year, a report appeared in the New England Journal of
 >     Medicine that predicted a decline in life expectancy in the United
 >     States as a direct result of obesity (Olshansky, et al. 2005).
 >
 >     Despite the assertions that obesity is causing our society great
 > harm,
 >     however, many scientists and activist groups have disputed the level
 >     of danger that it actually poses. Indeed, a recent analysis presented
 >     in the Journal of the American Medical Association (JAMA) by
 > Katherine
 >     Flegal of the CDC and her colleagues calls the severity of the
 > dangers
 >     of excess body fat into question, indicating that the number of
 >     overweight and obesity-related deaths is actually about 26,000about
 >     one fifteenth the earlier estimate of 400,000 (Flegal et al. 2005).
 >
 >     There is little argument about the fact that, as a nation, more of us
 >     are fatter than ever before; the disagreement lies in the effect that
 >     this has on our health. The campaign to convince us to lose weight
 >     gained much of its momentum in 2004; not only were there high-profile
 >     public health initiatives devoted to stopping the obesity epidemic,
 >     but the idea had pervaded popular culture as well. Movies like Morgan
 >     Spurlocks Super Size Me were the topic of many a discussion, and
 > there
 >     were regular news reports about the dangers of too much fat.
 >
 >     During this campaign, however, there were some notable dissenters.
 >     Paul Ernsberger, a professor of nutrition at Case Western Reserve
 >     University, has been doing research since the 1980s that led him to
 >     assert that obesity is not the cause of ill health but rather the
 >     effect of sedentary living and poor nutrition, which are the actual
 >     causes. Another prominent researcher, Steven Blair, director of the
 >     Cooper Institute of Aerobics Research in Dallas, Texas, has been an
 >     author on several studies indicating that the risks associated with
 >     obesity can be significantly reduced if one engages in regular
 >     physical activity, even if weight loss is not present. According to
 >     Blair, weight loss should not be ignored but a greater focus
 > should be
 >     placed on physical activity and good nutrition. Both Ernsberger and
 >     Blair indicated to me that they thought the new research by Flegal
 > and
 >     her colleagues provides a more accurate picture of the mortality risk
 >     associated with obesity.
 >
 >     [Obesity-Poster.jpg]
 >     Figure 1. This advertisement, paid for by the Center for Consumer
 >     Freedom (CCF), ran in magazines and newspapers across the country.
 > The
 >     ad was issued in response to the study in the Journal of the American
 >     Medical Association that found obesity-caused death rates had been
 >     exaggerated. However, CCF, an advocacy group for restaurants and food
 >     companies, has its own agenda.
 >
 >     While scientists like Ernsberger and Blair have been presenting their
 >     conclusions in the scientific forum, others have taken a more
 >     inflammatory approach. In his 2004 book, The Obesity Myth, Paul
 > Campos
 >     argues that the public health problem we have associated with obesity
 >     is a myth and further claims that our loathing of fat has damaged our
 >     culture (see Benjamin Radfords review on page 50). The most
 >     antagonistic group, however, is the Center for Consumer Freedom (CCF)
 >     (www.consumerfreedom.com), which implies that the obesity epidemic is
 >     a conspiracy between the pharmaceutical industries and the public
 >     health establishment to create a better market for weight-loss drugs.
 >     Numerous articles on the organizations Web site bash several of the
 >     most prominent obesity researchers who have disclosed financial ties
 >     to the pharmaceutical industries. Paul Ernsberger echoed this
 >     sentiment. He told me that the inflated mortality statistics were all
 >     based on the work of David Allison, a well-known pharmacoeconomics
 >     expert. These experts create cost-benefit analyses which are part of
 >     all drug applications to the FDA. These self-serving analyses
 > start by
 >     exaggerating as much as possible the cost to society of the
 > ailment to
 >     be treated (obesity in the case of weight-loss drugs). The risks
 >     associated with the new drug are severely underestimated, which
 >     results in an extremely favorable risk-benefit analysis, which is
 >     almost never realized once the drug is on the market. Experts who can
 >     produce highly favorable risk-benefit analyses are very much in
 >     demand, however.
 >
 >     The claims made by the CCF are given some credence by Ernsbergers
 >     corroboration; however, there is a noteworthy problem with their own
 >     objectivity. On their Web site they present themselves as a
 >     consumer-minded libertarian group that exists to promote personal
 >     responsibility and protect consumer choices. Upon closer examination,
 >     however, it becomes evident that the CCF is an advocacy group for
 >     restaurants and food companies, who have as much to gain by the
 > threat
 >     of obesity being a myth as the pharmaceutical industry does by the
 >     danger being dire.
 >
 >     It is clear that there are agenda-determined interests on both sides
 >     of the issue. Therefore, the best way to discern what is necessary
 > for
 >     good health is to shift our focus away from the sensational parts of
 >     the controversy and look at the science itself.
 >
 > Current Science and Obesity Risks
 >
 >     In their recent article, Katherine Flegal and her colleagues (2005)
 >     point out that the earlier mortality estimates were based on analyses
 >     that were methodologically flawed because in their calculations the
 >     authors used adjusted relative risks in an equation that was
 > developed
 >     for unadjusted relative risk. This, according to Flegals group, meant
 >     that the old estimates only partially accounted for confounding
 >     factors. The older estimates, furthermore did not account for
 >     variation by age in the relation of body weight to mortality, and did
 >     not include measures of uncertainty in the form of [standard errors]
 >     or confidence intervals. These authors also point out that the
 >     previous estimates relied on studies that had notable limitations:
 >     Four of six included only older data (two studies ended follow-up in
 >     the 1970s and two in the 1980s), three had only self-reported weight
 >     and height, three had data only from small geographic areas, and one
 >     study included only women. Only one data set, the National Health and
 >     Nutrition Examination Survey I, was nationally representative (Flegal
 >     et al. 2005). In their current investigation, Flegals group addressed
 >     this problem by using data only from nationally representative
 > samples
 >     with measured heights and weights. Further, they accounted for
 >     confounding variables and included standard errors for the estimates.
 >
 >     Obesity was determined in this analysis using each subjects body mass
 >     index, which is a simple height-to-weight ratio. A BMI of 18 to 24 is
 >     considered to be the normal weight, 2529 is considered overweight,
 > and
 >     30 and above is considered obese. The data from this study indicated
 >     that people who were underweight experienced 33,746 more deaths than
 >     normal-weight people, and that people who were overweight or obese
 >     experienced 25,814 more deaths than the normal-weight folks. This
 >     estimate is being reported in the popular media as being one-
 >     fifteenth the earlier estimate of 400,000. However, conflating the
 >     categories of overweight and obesity this way is misleading.
 >
 >     At first glance, it appears that underweight poses a bigger threat to
 >     our health than overweight and obesity, and that the earlier
 > estimates
 >     were profoundly exaggerated. However, in this study the people who
 > fit
 >     into the obese category actually experienced 111,909 excess deaths
 >     compared to normal-weight subjects. In contrast, those who were
 >     categorized as overweight experienced 86,094 fewer deaths than those
 >     who were normal weight. The figure of 25,815 is the difference
 > between
 >     the obesity deaths and the overweight survivals. In the original
 > study
 >     by David Allison and his colleagues (Allison et al. 1999) it is
 >     actually estimated that 280,000 deaths result from overweight and
 >     obesity and that 80 percent, or 224,000, of these deaths occurred in
 >     people who were in the obese category. However, the study by Mokdad
 >     and colleagues (2004), using the same methods developed by Allison et
 >     al., estimated 400,000 obesity-related deaths, and subsequently
 > fueled
 >     much of the recent fervor surrounding the obesity epidemic. In this
 >     study, no distinction was made between overweight and obesity and the
 >     authors failed to distinguish between obesity, physical inactivity,
 >     and poor diet. All of these variables were simply lumped together.
 >
 >     A few things become clearer after examining the data. First, it
 >     appears that our categories are mislabeled; being classified as
 >     overweight appears to give one an advantage (statistically, anyway)
 >     over those who are in the ideal weight range. [1] Moreover, it is
 >     inappropriate to consider overweight and obese as one group. Despite
 >     the current hype, the initial overestimation by Allison and his group
 >     was not as exaggerated as is being publicized; compared to that
 > study,
 >     the new estimate is actually about half of the old number.
 > Finally, it
 >     is apparent that many at the CDC were simply confirming their own
 >     biases when they accepted the estimate by Mokdad et al. The
 > categories
 >     in that studythat was, intriguingly, co-authored by CDC director
 > Julie
 >     Gerberding, which may provide some insight into why it was so readily
 >     acceptedwere far too broad to provide useful information. The fact
 >     that this flaw was ignored shows how easy it is to accept evidence
 >     that supports our preconceived notions or our political agendas.
 >
 >     There is another problem inherent in all of the above mortality
 >     estimates. They are based on epidemiological data that show
 >     correlation but leave us guessing as to causation. Various factors
 > are
 >     interrelated with increased mortalityobesity, inactivity, poor
 >     nutrition, smoking, etc. Yet, without carefully controlled
 >     experiments, it is hard to determine which factors causeand which are
 >     symptoms ofpoor health. This is a difficult limitation to overcome,
 >     however, because we cant recruit subjects and have them get fat to
 > see
 >     if they get sick and/or die sooner. Most institutional review boards
 >     would not approve that sort of research, and furthermore I cant
 >     imagine that there would be a large pool of subjects willing to
 >     participate. There are, however, observational data that were
 >     collected with fitness in mind, which help to clarify the picture
 >     somewhat.
 >
 >     In 1970 researchers at the Cooper Institute for Aerobics Research in
 >     Dallas, Texas, began to gather data for a longitudinal study that was
 >     called, pragmatically enough, the Aerobics Center Longitudinal Study
 >     (ACLS). This study looked at a variety of different variables to
 >     estimate the health risks and benefits of certain behaviors and
 >     lifestyle choices. What set this study apart from other large-scale
 >     observational studies, however, was that instead of relying on
 >     self-reporting for variables like exercise habits, they tested
 > fitness
 >     levels directly by way of a graded exercise test (GXT). A GXT
 > requires
 >     a person to walk on a treadmill as long as he or she can with
 >     increases in speed and incline at regular intervals. This is the most
 >     reliable way we know of to assess a persons physical fitness.
 >
 >     With an accurate measure of the subjects fitness levels, researchers
 >     at the Cooper Institute have been able to include fitness as a
 >     covariate with obesity. Analysis of the data obtained in the ACLS
 >     shows that there is a risk associated with obesity, but when you
 >     control for physical activity, much of that risk disappears
 > (Church et
 >     al. 2004; Katzmarzyk et al. 2004; Katzmarzyk et al. 2004; Lee et al.
 >     1999). One study showed that obese men who performed regular exercise
 >     had a lower risk of developing cardiovascular disease than lean men
 >     who were out of shape (Lee et al. 1999).
 >
 >     Steven Blair, who runs the Cooper Institute and was an author on all
 >     four of the above-mentioned studies, however, does not think obesity
 >     should be ignored. I do think obesity is a public health problem,
 >     although I also think that the primary cause of the obesity epidemic
 >     is a declining level of average daily energy expenditure . . . it
 > will
 >     be unfortunate if it is now assumed that we should ignore obesity. I
 >     do not think that the [health] risk of obesity is a myth, although it
 >     has been overestimated. Blair believes that a focus on good nutrition
 >     and increased physical activity rather than on weight loss will
 > better
 >     serve us.
 >
 >     In spite of the fact that there are virtually no controlled clinical
 >     trials examining the effects of obesity in people, we can make some
 >     inferences from animal research. Investigations performed by
 >     Ernsberger and his colleagues have shown that, over time, weight
 >     cycling (temporary weight loss followed by a regain of that weight,
 >     otherwise known as yo-yoing) in obese laboratory animals increases
 >     blood pressure, enlarges the heart, damages the kidney, increases
 >     abdominal fat deposits, and promotes further weight gain (Ernsberger
 >     and Koletsky 1993; Ernsberger et al. 1996; Ernsberger and Koletsky
 >     1999). This indicates that the yo-yo effect of crash dieting may be
 >     the cause of many of the problems we attribute to simply being fat.
 >
 >     Even though there is a health risk from being too fat, you can
 >     eliminate much of the potential risk by exercising. Moreover, it is
 >     probably a bad idea to jump from diet to diet given the negative
 >     consequences the yo-yo effect can have. According to another study
 >     published in JAMA, the risk of cardiovascular disease has declined
 >     across all BMI groups over the past forty years as the result of
 >     better drugs (Gregg et al. 2005).
 >
 >     None of this means, however, that we should simply abandon our
 >     attempts to maintain a healthy weight; obese people had twice the
 >     incidence of hypertension compared to lean people and, most
 >     significantly, there has been (according to the above study) a 55
 >     percent increase in diabetes [2] that corresponds to the increase in
 >     obesity. So while we are better at dealing with the problem once it
 >     occurs, it is still better to avoid developing the problem in the
 >     first place.
 >
 > Condemning the CDC
 >
 >     Whatever side of the argument you are on, it is apparent that many in
 >     the CDC acted irresponsibly. However, despite the fact that the
 >     initial, exaggerated estimate came from people at the CDC, we should
 >     keep in mind that so did the corrected number. While this can be
 >     frustrating to the casual observer, it is also a testament to the
 >     corrective power of the scientific method.
 >
 >     Science is about provisional truths that can be changed when evidence
 >     indicates that they should be. The fact that scientific
 > information is
 >     available to the public is its greatest strength. Most of us, for
 >     whatever reasonwhether its self-interest or self-delusiondont view
 > our
 >     own ideas as critically as we should. The fact that scientific ideas
 >     are available for all to see allows those who disagree to disprove
 >     them. This is what has happened at the CDC; the most current study
 > has
 >     addressed the flaws of the earlier studies. It is true that many of
 >     those in power at the CDC uncritically embraced the earlier estimates
 >     and overreacted, or worse simply accepted research that was flawed
 >     because it bolstered their agendas. But that failure lies with the
 >     people involved, not with the CDC as an institution or with the
 >     science itself.
 >
 >     The evidence still shows that morbid obesity is associated with an
 >     increased likelihood of developing disease and suffering from early
 >     mortality, but it also shows that those who are a few pounds
 >     overweight dont need to panic. Whats more, it is clear that everyone,
 >     fat or thin, will benefit from regular exercise regardless of whether
 >     they lose weight.
 >
 >     The lesson to be learned from this controversy is that rational
 >     moderation is in order. Disproving one extreme idea does not prove
 > the
 >     opposite extreme. As Steven Blair told me, It is time to focus our
 >     attention on the key behaviors of eating a healthful diet (plenty of
 >     fruits and veggies, a lot of whole grains, and not too much fat and
 >     alcohol) and being physically active every day.
 >
 > Notes
 >
 >     1. This is not the first time this has been shown. The following
 >     studies are also large-scale epidemiological studies that have found
 >     the overweight category is where the longest lifespan occurs: Waaler
 >     H.T. 1984. Height and weight and mortality: The Norwegian experience.
 >     Acta Medica Scandinavica Supplementum 679, 156; and Hirdes, J.,
 >     Forbes, W. 1992. The importance of social relationships,
 >     socieoeconomic status and health practices with respect to mortality
 >     in healthy Ontario males. Journal of Clinical Epidemiology 45:175182.
 >
 >     2. This is for both diagnosed and undiagnosed individuals.
 >
 > References
 >
 >     Allison, D.B., et al. 1999. Annual deaths attributable to obesity in
 >     the United States. Journal of the American Medical Association 282:
 >     153038.
 >
 >     Blair, Steven, and James Morrow, Jr. 2005. Comments on U.S. dietary
 >     guidelines. Journal of Physical Activity and Health 2: 137142.
 >
 >     Campos, Paul. 2004. The Obesity Myth. New York, New York: Gotham
 >     Books.
 >
 >     Church, T., et al. 2004. Exercise capacity and body composition as
 >     predictorof mortality among men with diabetes. Diabetes Care 27(1):
 >     8388.
 >
 >     Ernsberger, Paul, and Richard Koletsky. 1993. Biomedical rationale
 > for
 >     a wellness approach to obesity: An alternative to a focus on weight
 >     loss. Journal of Social Issues 55(2): 221259
 >
 >     Ernsberger, Paul, and Richard Koletsky. 1999. Weight cycling and
 >     mortality: support from animal studies. Journal of the American
 >     Medical Association 269: 1116.
 >
 >     Ernsberger P., et al. 1994. Refeeding hypertension in obese
 >     spontaneously hypertensive rats. Hypertension 24: 699705.
 >
 >     Ernsberger P., et al. 1996. Consequences of weight cycling in obese
 >     spontaneously hypertensive rats. American Journal of Physiology:
 >     Regulatory, Integrative and Comparative Physiology 270: R864R872.
 >
 >     Flegal, Katherine M., et al. 2000. Journal of the American Medical
 >     Association 288(14): 17231727.
 >
 >     Flegal, K., et al. 2005. Excess deaths associated with underweight,
 >     overweight, and obesity. Journal of the American Medical Association
 >     293(15): 186167.
 >
 >     Gregg, E., et al. 2005. Secular trends in cardiovascular disease risk
 >     factors according to body mass index in U.S. adults. Journal of the
 >     American Medical Association 293(15): 186874.
 >
 >     Hedley, A., et al. 2004. Prevalence of overweight and obesity
 > among US
 >     children, adolescents, and adults, 19992000. Journal of the American
 >     Medical Association 291: 28472850.
 >
 >     Heini, Adrian F., and Roland L. Weinsier. 1997. Divergent trends in
 >     obesity and fat intake patterns: The American paradox. Journal of the
 >     American Medical Association 102(3): 254264.
 >
 >     Katzmarzyk, Peter, et al. 2004. Metabolic syndrome, obesity, and
 >     mortality. Diabetes Care 28(2): 39197.
 >
 >     Katzmarzyk, Peter, Timothy Church, and Steven Blair. 2004.
 >     Cardiorespiratory fitness attenuates the effects of the metabolic
 >     syndrome on all-cause and cardiovascular disease mortality in men.
 >     Archives of Internal Medicine 164: 109297.
 >
 >     Lee, Chong Do, Steven Blair, and Andrew Jackson. 1999.
 >     Cardiorespiratory fitness, body composition, and all-cause and
 >     cardiovascular disease mortality in men. American Journal of Clinical
 >     Nutrition 69: 37380.
 >
 >     Mark, David. 2005. Deaths attributable to obesity. Journal of the
 >     American Medical Association 293(15): 191819.
 >
 >     Mokdad, A.H., et al. 2004. Actual causes of death in the United
 >     States. Journal of the American Medical Association 291: 123845.
 >
 >     Olshansky, S, Jay., et al. 2005. A potential decline in life
 >     expectancy in the United States in the 21st century. New England
 >     Journal of Medicine 352(11): 113845.
 >
 > About the Author
 >
 >     Patrick Johnson is a biology instructor at Washtenaw Community
 > College
 >     in southeast Michigan and a clinical exercise physiologist who writes
 >     frequently about health, nutrition, and fitness claims. He lives with
 >     his wife and his eight-year-old son. E-mail: johnsonp @wccnet.edu.
 > _______________________________________________
 > paleopsych mailing list
 > paleopsych at paleopsych.org
 > http://lists.paleopsych.org/mailman/listinfo/paleopsych
 >