This is a long post. Topics:

* obesity and what public health policy has to say about it

* correlation and causation fallacies

* incentives

* free will

A friend sends me this article from the New England Journal of Medicine. (I’m pretty sure it’s not behind a paywall, but I could be wrong. If you have troubles getting to the article let me know and I’ll see about sending you a copy.) It’s about obesity and what ought to be done to stop (reduce) it.

IOM Obesity Report. It is the primary topic of discussion of the NEJM article (above link) and I will refer to it.

This is going to be a very pedantic post. More so than usual. You’ve been warned. Also, I’m going to offend. Guaranteed. More so than usual. You’ve been warned. Stop reading if you can take it.

Before we even get into the article, let me point out this (because it forms the foundation of many of my later criticisms, both logical criticisms and normative ones): obesity is – by definition – a personal problem that nobody other than the afflicted individual need care about. The only way this is not true is if the cost of obesity is pushed onto others, say, through public health care or laws requiring emergency rooms to treat patients, or other legal means. (Of course, obesity IS pushed onto others through public health policy, so my prior statement might sound a little absurd. It’s not: that obesity is public and not private is entirely an arbitrary choice of the society and government in which we live.) I contrast obesity, in this light, with a transmittable bacteria, disease, or virus. The reason this distinction is important is because the degree to which society does, in fact, bear the cost of dealing with obesity (through any number of publicly financed health care initiatives, including taxes that are paid to support federal agencies that engage in public awareness campaigns) is a choice. Secondly, and relatedly, the degree to which society does, in fact, bear the cost of dealing with obesity acts as an incentive to be obese. Or, put conversely, the degree to which public health initiatives exist and help obese people reduces the cost (both monetary and health) of being obese. These are powerful incentives that absolutely cannot be overlooked in assessing how to solve the obesity “problem”. Indeed, the extent to which obesity is a problem at all can only really be understood from the perspective of: society must pay health care costs collectively and as such must reign in the expense of dealing with those above the average. If we simply adopt a policy of: I pay my own expenses, then the problem entirely vanishes (for society, not for the obese individual). But lest this sound dismissive, let me point out that this is no different than if we (as a society) were to declare, for example, that not owning a iPad is bad. If not owning an iPad is bad, then this country has a big iPad-lessness problem that must be solved and can be solved through collective action. And, ultimately, the reason this is important is that because with these particular incentives in place it is very hard (impossible?) to understand the “true” motivations of the people that are or aren’t obese.

Preamble finished, back to the article…

One is immediately (second paragraph) struck by the non-scientific nature of the review. Here is the article:

“The groundbreaking report and accompanying HBO documentary, “The Weight of the Nation,” present a forceful case”

Hmm…a “forceful case”. This is clearly rhetoric and not science we’re talking about. I’m not saying that’s a priori bad. It’s not. Nor is there any deception on the part of the journal, as the article is published in the “Perspectives” section (akin to a newspaper OpEd section). It’s just important to keep this in mind. Scientists are people and often they are not scientific. This article has already raised this red flag. Often one comes to a publication like the NEJM thinking that it’s all (good) science. Well, often it’s not (science, nor good science).

Here is a statement by the authors that is simply not true:

“For example, the report details accumulating evidence that the obesity epidemic has been driven by a complex interaction of changing factors in several critical environments — our schools, workplaces, communities, media, and food and beverage systems — rather than by individual choices.”

If you go into the IOM report, you find two full pages (49, 51, with graphs/charts surrounding and in between) that detail exactly what the problem is: eating too much food and exercising too little. Here is the key statement (very clearly made, I might add):

“Trends illustrate that, for both children and adolescents and adults, require-ments for physical activity have decreased while inducements to be inactive have increased, and food availability and consumption have increased.” (page 51, IOM report)

So now I’m a little peeved at the authors: they’re lying.

Now, granted, the NEJM authors are more focused on the solutions than the underlying problem, but I believe that in this case ignoring, getting incorrect, or flat-out lying about the underlying problem is detrimental to solving the problem (if you concede that there is a problem). Why?

Because misdiagnosing the problem leads to all sorts of stupid recommendations. Here’s one of my favs: the notion that access to healthy food (or, to “bad” food) is part of the solution (or problem). The NEJM authors allude to this as a cause with this veiled statement:

“Only 18% of Americans identify external factors (exposure to junk food, lack of safe places for children to play, and limited availability of healthy foods in some neighborhoods) as the biggest causes of childhood obesity”

There is clearly (dispute me on this if you want) an implication of correlation and causation. But while I concede that there is probably correlation (obese people are probably more surrounded by “bad” food than by good food), that does not mean that the solution lies in remedying this “cause”. Why not? Because it’s probably not a cause! Here’s the experiment that you have to do: give obese people access to better food choices. Do they take the better choice or the bad choice? Evidence (here) suggests that they take the bad choice.

On the exercise side of the equation, the IOM offers similar anti-causation-correlation evidence:

“Table 2-3 shows that, although both adults and children and adolescents reported increased leisure-time physical activity over time, recommended levels
of physical activity were not being achieved. A barrier to engaging in physical activity that is often cited is the lack of time; at least for adults, however, the availability of leisure time has increased over time.” (IOM, page 49)

We have more free time than ever and we’re not exercising! It can’t get much simpler than this!

Now, again, one can make the true claim against me that I’m not really addressing what the NEJM authors write about: effective public policy approaches to combat obesity. I’m not addressing that thesis at all. What I’m doing is combating the legitimacy of that as a thesis. From a public policy perspective do we really want to try to help people who so clearly do not want to help themselves? They reject good food choices when they have them and they sit on the butts when presented with Wii Fit. Is this an audience that wants to be helped? Is this an audience that should be helped? And I would go so far as to ask: how condescending is it to try to force this audience to be helped? (Answer: very.) I would wager (with absolutely no supporting evidence) that buying iPads for everybody would yield a higher rate of return than combating obesity as vigorously as the authors might like to. Buying iPads would certainly be less pretentious.

Finally, and to underline all of the above, one absolutely cannot dismiss the fact that the authors are in a position to benefit greatly from any implementation of the recommendations in the article or the cited literature. They are public health policy “experts”. Who will receive money if their recommendations are implemented? Public health policy experts. This is akin to your auto mechanic saying that you should bring your car in for repair every week because of X. It does not matter what X is or if X is legitimate or not; your mechanic (or some other mechanic…the industry as a whole) will benefit from this recommendation. The wolf may or may not wear sheep’s clothing but it is something of which we must always be wary.

Happy to hear any and all feedback on this topic, in particular, if anybody has references to studies that show that people DO, in fact, chose healthy foods or exercise habits when presented with these options in an environment that previously lacked them I would love to see that. I am eager to be swayed on this.

-JD Cross

 PS: another absurd thing that the NEJM authors write (that’s not really that related to the above, but it’s just eating at me…):
“the report details accumulating evidence that the obesity epidemic has been driven by a complex interaction of changing factors in several critical environments”
Ok, so is the interaction complex or not? If it’s complex then that probably means that it’s hard (like, really hard) to understand. That probably means that it’s not understood. If it’s not complex then stop patting yourself on the back!
PPS: All this complaining has made me think that I should throw out my own solution to the obesity “problem”. Here it is: take whatever money that you might spend on all this public policy (read: pretentious, arrogant) social engineering and dump it instead into research into things like drugs that will do things like stimulate metabolism, prevent caloric uptake, etc. This solution (assuming that a drug or drugs came out of the process) would completely remove the elitist, social engineering, holier than thou aspect of the public health policy issue above. With the drug, individuals could choose to take it or not – no need for top-down, iron fisted control.