smoke 'em if you got 'em

Bee at Backreaction points to a study by Dutch scientists estimating the lifetime health costs caused by by obesity and smoking.

The outcome was that longlivety eventually turns out to be more expensive than living unhealthy and dying young:

"Although effective obesity prevention leads to a decrease in costs of obesity-related diseases, this decrease is offset by cost increases due to diseases unrelated to obesity in life-years gained. Obesity prevention may be an important and cost-effective way of improving public health, but it is not a cure for increasing health expenditures."

So far, this is one of those standard counterintuitive medico-science stories that always hits the news and gets people talking. Somehow, every smoker will see it and say, "see? I'm not the burden on society, you health nazis!" But what's interesting is how the researchers retrieved the result:

not through data analysis but by modeling three hypothetical groups: Obese non-smokers (BMI > 30), non-obese smokers, and non-obese non-smokers. These were then evaluated with regard to their respective statistical probability for certain illnesses from the age of 20 until the time when the model predicted their death. It turns out, though the highest yearly costs were incurred by the smokers in their older ages they were in total less costly since they died early, whereas the expenses for the non-obese non-smokers in their late years added up and the total lifetime health spending was larger. The researchers estimated the average cost for health treatments for the smokers to 220,000 EUR, for the obese non-smokers to 250,000 EUR and for the the non-obese non-smokers to 281,000 EUR.

But will such a model really work? The first phrase that really worries me is, "certain illnesses." If the scientists have cherry-picked expensive chronic diseases, then possibly the method makes sense (since otherwise healthy individuals might live longer with their treatments). But if mortality-increasing diseases are being modeled, a key piece of the causal picture has been removed (by making incidence and longevity independent parameters).

In their methodology, the authors explain that "risk factors were linked to 22 obesity- and/or smoking-related chronic diseases through relative risks of disease incidence for each risk factor level, to model the chain leading from risk factor to disease to death." [Incidentally, they cite another paper as the basis for this conclusion. I've not looked at it.] Possibly I'm misinterpreting, but this seems to support my initial read: a healthier person with a chronic smoking- or obesity-related disease will live with it longer, so much so that their medical costs outweigh the savings from their cohort's overall lowered incidence of said disease. Put another way, if you have one of these diseases, your risk of death increases significantly if you are a smoker or are obese (since, for healthier individuals, these chronic diseases are more manageable).

It should be noted, the study reflects both risk factors and medical costs in the Netherlands. The picture certainly looks different in the United States or Canada.

Another interesting finding is that obese people have higher health costs between 20 and 56, at which point smokers surpass them. Later, when all the obese smokers have died, the health-nuts are still plugging along, a drain on the system. In other words, otherwise UNhealthy people are subsidizing the treatment of otherwise healthy people with these 22 diseases.

Does this mean insurance companies should do away with healthy-living rebates? Or that they will mail you cigarette starter packs and "Red Meat Digest" upon your retirement?