Let us begin the discussion.
1. First we pose the four questions we need to answer. This we show below.
The issue of what causes Type 2 Diabetes has been looked at by many and generally we all know it is a carbohydrate disorder, along with some lipid and protein issues as well, by as Banting and Best knew from the then clinical practice on Type 1 Diabetes, controlling carbs, near starvation, kept Type 1 patients alive, barely, but Type 2 patients could actually revert to normal if the BMI was brought below 22.5 early on in the disease.
2. If we can answer the above questions then we can put them in a simple economic model for costing out the overall societal costs for Type 2 Diabetes. We summarize that below.
So we can see what the incidence, and in turn prevalence is for T2 Diabetes and then add it to the model. Remember incidence is new cases and prevalence is total so we must add all the above up. Also one must remember that T2 sequellae are long lasting. Unlike lung cancer, where the patient frequently dies quickly, T2 sequallae patients may linger for years, well into the Medicare period as well.
3. Obesity causes T2 Diabetes.
There are multiple papers and books with the latest summary being in Science in 2009 by Lazar. There may still be some rough edges to the analysis, but it is compelling, more so than almost all other studies. There is also the book by Mantzoros on Obesity and Diabetes which provides substantial data through 2005.
One may then ask what causes obesity, for the most part it is simply excess calories, but we will return to that. The answer of the nexus between obesity and T2 Diabetes is unquestionable, the details are a work in progress, and the literature is extensive. Clinically any physician in day to day practice dealing with T2 Diabetes is also confronting obesity. It has been that way for years.
4. The prevalence of T2 Diabetes is increasing significantly. We show the data below. Clearly the growth from 1070 shows the main driver between T2 Diabetes and obesity.
5. If one were to plot the cost of health care as a % of GDP versus the % prevalence of obesity in the US one obtains the curve below.
This curve, albeit statistical, and separate from the studies under 3 above, shows a strong correlation. It shows costs due to obesity taking over the health care expenditures. It shows that there is a strong statistical correlation between obesity and the explosion in health care costs. We will detail that shortly.
6. So what does T2 Diabetes lead to. The next set of questions we posed under 1 above. Well we all know the answer, cardiovascular, nephropathy, neuropathy, retinopathy, and the list goes on!
The problem with these is that they last a long time and cost lots of money. Take a stroke, that can have costs which last decades! Again dramatically unlike smoking and many cancers. Cancers are either cured or they kill you. For the most part we have limited success turning cancer into a chronic disease. CML may become one of the first.
7. Here is some of the most recent data on the sequellae, their incidence, prevalence, and treatment options.
This is quite costly.
8. Looking forward we now have to worry about the lifetime risks of T2 Diabetes. This is shown below.
This means that one third of the people born in 2000 will develop T2 Diabetes! Does anyone really know what that means. It is not the cancer problem, we may have actually solved that by the time this tsunami hits.
9. These are some of the sequellae that this growth will cause.
10. And these are some of the current stats on the disease as we see them today.
11. We now want to goes back to item 2 above to continue to fill in our cost model. This can be followed as below:
We now need to get data for the costs of the sequellae and better stats on their occurence. We do that as follows.
12. The following is a great study which benchmarked all of these costs. It is one of dozens but it is the one which I have come to rely upon.
The study we well done and should be studied.
13. The study performs the steps as we show below. We reached the answer that in 2010 it costs $300 B. That can be demonstrated in the next point.
14. The details of the cost model using the 2007 study yield the following data:
Here we took data for 2008, demographics and prevalence, and then used the study data from Brandle et al to modify the data in the above table. Then we used the cost increase data from 2008 to 2010 and the increased prevalence from 2008 t0 2010 and the increased population from 2008 to 2010 and voila we get $300 B! That is 2X the CDC numbers from 2007. But we argue that the numbers must (i) reflect the up tick as we just described it, and that (ii) the CDC numbers should be adjusted to reflect the Brandle methodology.
15. The following is a summary of my argument from yesterday and the day before regarding costs on a per carb basis.
Here we have laid out the logic and the referring source materials. Is this the definitive study, not yet, it will continue to evolve.
Mankiw responded in his blog yesterday as follows:
When writing my article, I contacted several prominent health economists to ask whether a complete accounting of both budgetary costs and benefits has been done for obesity, as has been done for smoking.
Yes the definitive study has not been completed, most likely will never be done. Yet this analysis and the ones we refer to provide a reasonable basis for policy formulation. The approach here is still one of a "back of the envelope" albeit in my book I have provided adequate detail, perhaps not of the PhD level, but at a level that a rational business person would use to make an informed decision. I would assume that such would apply to politicians, but alas I left DC 30 plus years ago vowing never to return.
This explanation took me a little less than 3 hours, almost half working on getting it on Blogger. Yet I spent almost a year working on the problem, having spent 25 plus years thinking and working in the area. Is this analysis fully buttoned up, no, there are loose ends, but is is a map to the solution, at least one map. Enjoy the journey, but this is a real problem.
One final note, unlike the Pigou tax and the gasoline issue, here with the carb tax there is an alternative. You can stop eating, lose the weight and return to health. At that point you are both health and not paying the tax. No such alternative exists for those who drive.
Update: Just noticed that the National Soft Drink Association hit my blog from yesterday this morning. Just to be fair, my position is not against soda it is carbs in general. As someone who took his BMI from 30.5 to well below 25 eight years ago, and kept it there, I am like an ex-smoker or a religious convert, I am a zealot advocate since I see what works. I have not had a soda in eight years, I remember the day, and I remember it was root beer, I really love root beer, but as a recovering carboholic I remain staunchly against carbs!
Posts
