A report released this week puts a surprisingly high figure on the societal cost of obesity in the US: $1.72 trillion annually, or 9.3 percent of GDP. By contrast, the current CDC estimates are in the region of $150 billion, less than one tenth as high.
By far the biggest chunk of that $1.72 trillion is the $1.24 trillion chunk attributed to the “indirect” costs of obesity: the “work absences, lost wages, and reduced economic productivity for the individuals suffering from the conditions and their family caregivers,” the report explains. That is, the bulk comes from costs other than healthcare spending. The estimate for healthcare spending—$480.7 billion annually—is somewhat higher than a range of estimates in reviews of the literature, which hover around $150 to $300 billion, but are still on the same scale.
Estimates like these can vary substantially because of the different methods used in calculating them. For example, a review from 2017 catalogues the different obesity-related diseases that were included in various studies across different countries; respiratory disorders and musculo-skeletal disorders make an appearance in some but not others. A 2016 meta-analysis describes a similarly wide range in how medical costs are calculated.
How much cancer is obesity-related?
The new report was published by the Milken Institute, a think tank that aims to “increase global prosperity by… widen[ing] access to capital, creat[ing] jobs, and improv[ing] health.” That means that, while the report leans heavily on the academic literature, it hasnt been scrutinized through the process of scientific peer review and doesnt have the same obligations as a scientific paper to publish the details of its workings.
To reach their estimate, public health economist Hugh Waters and health research analyst Marlon Graf gathered data on a range of chronic health conditions. That range included obvious culprits like stroke and Type 2 diabetes as well as conditions like chronic back pain and liver cancer. They used CDC data to establish how common these diseases are in the US.
Next, they calculated how much of this prevalence could be attributed to obesity. Obviously, it would be insane to consider every case of liver cancer to be the result of obesity, even though obesity is linked to a higher risk of various cancers. So, for instance, to work out what proportion of the total prevalence of liver cancer could be attributed to obesity, Waters and Graf looked at how much higher the risk of liver cancer is for someone with obesity compared to someone without obesity, and they extrapolated that to the overall prevalence numbers.
The data on costs came from a few different sources—a national survey that includes a household survey, along with data from medical insurance, the CDC, and various academic sources. These sources cover both the direct medical costs and the indirect costs of each condition; for instance, one paper estimates the total cost of chronic back pain in the US.
Combined with the figures on disease prevalence, this allowed Waters and Graf to reach their new estimate of $1.72 trillion annually.
Pinning economic value on humans
This method casts the net widely, drawing a large circle around what counts as an obesity-related cost—a set of choices made clear by the fact that the estimate is so high compared to others. But it illustrates just how much different data, different assumptions, and different methods can affect the overall results.
There are still other factors that the report could have taken into account when casting its net. For example, reducing the prevalence of fatal diseases leads to a society with longer lifespans—and that brings with it the associated costs, both direct and indirect, of an aging society. Thats an incredibly complex relationship: diseases have varying fatality rates and different costs of treatment, making it impossible to guess whether taking this into account would have increased or decreased the estimate of healthcare costs.
Another complication pops up, too: additional lifespan doesnt only bring cost—it also brings economic gain. “There are, in fact, quantitative methods used by health economists to place a dollar value on additional years of healthy life gained through health promotion and health care,” says Waters. “These types of studies usually value a life year gained as worth about $50,000—and the value of a life saved as between $2 million and $3 million.” If their estimate had included this, he adds, “the total costs of obesity would have been much higher.”
Adding these factors into the equation would introduce new layers of complexity, given the constantly shifting sands of both illness prevalence and how much various conditions costs to treat. Then there are complexities in how healthcare costs change with age: people close to death have higher healthcare costs, regardless of what age they are. And of course, these calculations would all play out differently in countries with other healthcare systems.
Research that compares the costs of fatal diseases to the costs of longer lifespans is often conducted with the starting assumption that “people living healthier and longer lives is an inherently positive social goal,” Waters says—this research isnt arguing that fatal diseases are a good thing for society, just that the costs should be understood properly.
But the argument has been made elsewhere: Waters points to a tobacco industry-funded study making the case that “smoking was good for the public finances of the Czech Republic, precisely because people were dying younger and not using all of their state pensions or state-funded health care.” Philip Morris, he adds, did not come out of that looking very good.
Any attempt to turn human lives into numbers, however necessary, appears in the middle of a fraught conversation on obesity, with wildly varying arguments all resting on the same set of facts. Although the costs of obesity are used to build an economic case for investing in public health, the cost of obesity is also used to justify and express prejudice against people with obesity. Because of this tug of war, precision becomes all the more essential—and that precision is clearly very difficult to achieve.
[contf] [contfnew] 
Ars Technica
[contfnewc] [contfnewc]
