...defined as a threat to societal values and interests: its nature is present in a stylized and stereotypical fashion by the mass media; the moral barricades are manned by editors, bishops, politicians and other right-thinking people; socially accredited expert pronounce their diagnoses and solutions; ways of coping are evolved or (more often) resorted to; the condition then disappears, submerges or deteriorate and becomes more visible. Sometimes the object of the panic is quite novel and at other times it is something which has been in existence long enough, but suddenly appears in the limelight. Sometimes the panic passes over and is forgotten, except in folklore and collective memory; at other times it has more serious and long-lasting repercussions and might produce such changes as those in legal and social policy or even in the way the society conceives itself. (from Society and Culture Bundle RC: Folk Devils and Moral Panics, by Stanley Cohen, p.1 - Routledge Classical edition, 2011)
In the discussion we had in response to my post on January 9, 2012, "The Protestant Health Ethic - Clinical Calvinism?" the question of risk was posed. Specifically, three kinds of risk that are commonly asserted regarding the "obesity epidemic": financial, military and social. As I promised during that discussion, I'd like to address each of these risks and make a case that instead of hard evidence, these three risks are the basis for creating a moral panic around obesity. I certainly will not be the first person to do this. For longer treatises see The Obesity Myth by Paul Campos, Fat Politics by J. Eric Oliver and Rethinking Thin by Gina Kolata. But I think there are some succinct points that can be made here about these risks. However, it is obvious that this will not be a comprehensive treatment. I mean only to outline arguments, not fully develop them, though I hope discussions will be useful. I will be breaking up these risk over three blog posts: cost/benefit analyses, military and social.
FINANCIAL: COST/BENEFIT ANALYSES
FINANCIAL: COST/BENEFIT ANALYSES
The most quoted study of late is a 2009 study published in Health Affairs ("Annual Medical Spending Attributable to Obesity," Finkelstein, et.al, October 2009), so I will use this study to demonstrate my problem with such estimates. I basically have four major objections to this study and others like it.
First, defining "cost" is problematic in a for-profit medical system. This criticism can be levied against almost any cost/benefit analysis regarding health care in the United States. A medical "cost" in a profit-making medical system is almost a dangling signifier. A cost is really a price in this situation. Do we include profits? advertising? high salaries? The Finkelstein study looks at expenditures. This means what was charged and paid by a funding source (Medicare, Medicaid, private insurance and out-of-pocket). Calling such a thing a cost is misleading. In our current system, pharmaceutical companies and HMOs post double-digit profits. The health care industry spent an estimated $1M a day on lobbying Congress during the healthcare debates in 2009. Many of these companies have astronomical advertising budgets that include taking doctors on Caribbean trips for "seminars." Patents give companies monopolies on medications and equipment. Some estimate that an extra 25% is spent in doctors' offices and hospitals to just administrate the complex procedures needed to receive payments from HMOs and insurance companies. Presenting a cost/benefit analysis without looking at this extra baggage is highly misleading. There should be at least some allowance for this overhead to ensure that the increase in costs isn't accounted for by an increase in the extras.
Second, actual collection of data is problematic. Specifically, the Finkelstein study uses Medicare and Medicaid data as well as some private funders. Medicare is more accurately kept than private funds (insurance companies protect data as proprietary business information), but recipients of Medicare are not typical of the general population. Finkelstein, et.al. admit that they are limited by their sources of data (p. w830), citing, among other problems, having to estimate costs of care while institutionalized. Since older adults who receive Medicare are also more likely to be hospitalize, this less-than-perfect source of data begins to complicate things quite a bit.
Third, like many other such studies, the Finkelstein estimates made no distinction regarding exactly what treatment was being given. They dismiss weight loss treatments as not being "widespread." They do not look at the weight histories of any of the people in the studies, even though weight-cycling has been shown to increase health risks. They do not look at aging, socioeconomic level, stigma, discrimination, stress, physical activity or access to health case (which delays treatment, making early stage intervention less likely). These confounding factors offer alternative explanations to increases in care and will not get better with weight loss. The implication of this study as presented is that making larger people smaller would make these costs go away. But nothing in this study demonstrates this.
Finally, one has to ask what gets defined as an "obesity-related" problem. Some studies before the Finkelstein study were calling all incidences of heart disease, diabetes, high blood pressure and metabolic disorders "obesity-related" despite the BMI of the person. Smaller people suffer from these conditions. The Finkelstein study looks at BMI but makes no distinction as to where the medical costs are "obesity-related" conditions. The implication is that weight loss would lower these costs, but without data regarding exactly what the expenditures were for, it is a big leap to assume making larger people smaller would lower costs.
An alternative reading of the data might suggest that the war on obesity is the source of increase costs rather than higher BMI itself. The past 16 years, since the War on Obesity was declared in 1996, we have seen a concerted public health push to treat larger people, including encouraging doctors to be more aggressive with treatments. Weight loss surgery is one of the fastest growing treatments, costing a great deal of money. Medicare, Medicaid and many private insurances have added coverage of such surgeries within the time frame of the study. Complications and deaths from this surgery are often reported as "obesity-related" conditions rather than surgery-related. Screening for metabolic disorders are encouraged and redefinitions of thresholds have led to many more people getting "treatment" than previously.
In other words, the panic itself may have increased costs:
Because BMI is considered a risk factor for many diseases, obese persons are automatically relegated to greater testing and treatment, which means that positing BMI as a risk factor results in increased costs, regardless of whether BMI itself is problematic. Yet using BMI as a proxy for health may be more costly than addressing health directly. ... The weight bias inherent in BMI profiling may actually result in higher costs and sicker people. (Bacon & Aphramor http://www.nutritionj.com/content/10/1/9)
Costs studies are really expenditure estimates. As with all estimates, the initial conditions of the data and how the data is used determine the end result. Finkelstein works for RTI, a consulting firm that lists a huge part of the medical-pharmaceutical-industrial complex as "clients." This calls into question the biases contained in their estimate. The fact that the CDC would use this as their basis for epidemiological assessment of costs on their website rather than making an independent assessment of their own, speaks to the difficulties of funding a more objective study. It may be cynical of me, but I would guess that a great deal of pressure is put on the government not to really find answers, especially answers that might make things like bariatric surgery not seem so cost effective. It is not surprising that one of the largest sources of the $147B meme was Allergan, the makers of lap band devices. They have quoted this estimate every chance they have and news media outlets have not questioned the original study or Allergan's profit motive when it is quoted.
"Studies have shown" are magic words in our society. Whatever words follow are treated as truth by the majority of people who hear them. The magic is that these words bestow credibility instead of skepticism. The proper response scientifically is to question, but those who profit from science prefer that the listener be "wowed" by data rather than inquisitive.
My greatest objection to such cost/benefit analysis, however, is on ethical grounds rather than scientific: Such estimates are almost always veiled attempts to vilify one portion of the population, especially when such estimates single out physical characteristics of people rather than incidences of disease or medical conditions. A cost analysis of diabetes is ethically different from a cost analysis of obesity because the former is based on diagnoses and treatment of a medical condition and the latter is based upon a demographic, BMI (weight and height). We cannot ethically decide that some people cost more than others.
I once heard a medical anthropologist discuss the US's health rankings in the world, dismissing our poor showing as a matter of poor people and minority groups bringing down the averages. In other words, because wealthy white people were as healthy as Canadians and Europeans, it was not important that we ranked high in infant mortality and low in life expectancy compared to other western countries. Poor Americans and minority group Americans are still Americans. You cannot data mine them out to make you feel better. This is prejudice, not science. The underpinnings of "costs of obesity" estimates are just as weak and questionable.