Bob, quite rightly, raises questions about the accuracy of prevalence data.

Let's work through the process by which this data is collected and used, looking at where inaccuracies can occur.

We start with the disease. There is a WHO classification of diseases [1]. However, errors can still be made:
- misdiagnosis is high in PD;
- timing of diagnosis is varied.

There is, it seems to me, to be a poor definition of when PD begins. Is it when the neuronal damage begins, or when the symptoms begin, or when the symptoms are noticed? I suspect that a perfect biomarker would identify the on-set of the disease at least 10 years earlier than the diagnosis by normal methods, leading to an artifical reporting of a doubling in the prevalence of PD.

We go on to look at the patient. For a case to be reported the patient needs to go to the doctor. In countries with poor health systems this is likely to happen later, if at all, than in countries with good health care systems. Even in countries with generally good systems, there may be groups of people who are disenfranchised on the basis of cost, or whatever, from seeing a doctor. Even when the patient has access to good healthcare there are differences in the timing of people's first presentation.

We still have the problem of whether the data is collated. It amazes me that Parkinson's is not everywhere a notifiable disease in which every case needs to be registered centrally.

Where doctors' records are not available epidemiologists have a number of alternative methods, these include:
1. they can do door to door surveys. This is only as good as people's preparedness to answer the door and answer the questions correctly.
2. they can use records of the prescription of certain drugs which probably signify a case of PD. (This is the method I adopted with my maps which estimate the geographical variations in the prevalence of PD in England). This is only as good as the extent to which a set of drugs differentiates PD from other diseases, such as restless legs syndrome.
3. cause of death records. These are only as good as the extent to which PD is mentioned.
4. surveying on the internet. But, this is only as good as the extent to which the people who respond are typical of the general population.
5. completely indirect approaches, such as internet search engine query analysis.

Let's assume we have the raw data. It's natural to want to make comparisons between places, between genders etc.. For an analysis to be meaningful we need to compare like with like.

The main, but not only, adjustment that we need to make is for age. Since the incidence of PD increase rapidly with age, comparing two populations with different age profiles would give misleading results. To get around this, adjustments are made which take into account a standard age distribution. See WHO [2]. This is still not the end of the matter. The age profile changes over time. For instance, it is likely that in 20 years time we will need to increase the weighting of 100+ people. The age profile you use depends on the question you are trying to answer: an international distribution makes sense if you are making comparisons between countries; but if you are making comparisons between parts of one country, a national comparison makes sense.

Once you accept the use of age standardization it follows that you must accept the normalization of other parameters such as gender and race. Unfortuately, the extent to which you can do this is limited by the number of cases. If you drill down too far, the number of cases in each category gets so small that statistical noise tales over.

This post has been about problems with epidemiology. That said, in my opinion, it remains one of the most likely routes to finding a cure for PD. From the point of view of those of us without access to electron microscopes, MRI scanners, (literal) lab rats etc., it gives a viable way in which we can try to find the cause of the disease.

References

[1] http://apps.who.int/classifications/...browse/2010/en

[2] http://www.who.int/healthinfo/paper31.pdf

John

Does anyone have experience with eating fresh turmeric for PD? I've recently been purchasing fresh turmeric from a local Asian supermarket and have been putting a couple pieces of the roots in a blender in the morning with frozen berries and juice to make a smoothie. Any one know of what quantity should be taken daily? Also, I'm making the assumption that fresh turmeric is better than dried curcumin or turmeric powder - can anyone validate this?

i'm no expert but i think you get a lot more active ingredient, cucurmin, from processed root, most of the fresh turmeric is water and fiber. plus who grows the root?
there are supposedly bioavailability issues too.

iherb sells NOW BIO-CURCUMIN PHYTOSOME and there is a similar product from JARROW
Jarrow Formulas, Curcumin Phytosome, 500 mg, 60 Capsules

The problem with curcumin, even with its extensive medicinal history, is that it doesn't absorb well when taken orally and very little crosses the BBB. Lawrence Helson, a reknowned oncologist (one of the discoverers of Taxol) is running a company now called SignPath Pharma. They are working of three versions of curcumin that will pass through the BBB. One, in conjuction with MD Anderson, is a liposomal curcumin (basically injected into a lipid). They just successfully completed a Phase I. Another, in conjuction with Johns Hopkins, is using Polymeric nanotechnology to make a nanocurcumin. Although this research is currently in cancer centers, Dr. Helson has stated that it could be used, if successful, for a host of neurological diseases, including PD.

My turmeric root that was put into a small container with potting mix in the kitchen in March FINALLY is growing and was transferred to a 5 gallon pot in the greenhouse. Excited and glad that I didn't give up on it! We shall see in a few months whether the fresh root is better than taking capsules of dried root! My two licorice plants are growing well also...so next year we may have some new experiments going on at Bean Acres.