The CDC has produced a table showing, state by state, the number of deaths reported to be, at least in part, due to Parkinson's in the US in 2010 [1]. The data is shown in the Appendix. The rate values are per 100,000 of population.

Using this data and the R programming language map package, I've made a map showing the age adjusted PD mortality rates across the US.

PDmortality2010map.png

Parkinson's is well known to be under-reported as a cause of death. But, this is not too much of a problem if we confine ourselves to relative differences, i.e why does state A have 20% higher figures than state B? (Of course, inconsistencies between the reporting conventions of states will affect the usefulness of the numbers.)

It is useful to compare the CDC figures with those in the paper by Willis et al. [2], which has already been mentioned in this thread. The Willis paper reports prevalence and incidence figures, not mortality rates. However, as a rough measure, for a disease like Parkinson's one would expect the annual mortality rate to be similar to the incidence rate. This is because the number of people joining the pool of PwP should be similar to the number leaving it.

I am surprised by the apparent only limited similarity between the two distributions.

Some explanation can be put down to:
- both sets of data are age standardized, but the Willis results are also race standardized.
- the Willis paper is based on Medicare records, which affects the age distribution of its data.
- the Willis paper is on a county by county basis, perhaps a whole state is affected by just a few high density counties.
- for the smaller states the numbers are low, e.g. Delaware reported 57 PD related deaths, suggesting that there could be significant differences from year to year.

Why does this matter? It matters because spatial differences in the rate of Parkinson's offer a good way of identifying environmental factors in the etiology of PD.

References

[1] http://www.cdc.gov/nchs/data/dvs/dea...10_release.pdf

[2] "Geographic and Ethnic Variation in Parkinson Disease: A Population-Based Study of US Medicare Beneficiaries"
Allison Wright Willis, Bradley A. Evanoff, Min Lian, Susan R. Criswell, and Brad A. Racette
Neuroepidemiology. 2010 April; 34(3); 143-151.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2865395/

John

Appendix: CDC data for 2010 PD mortalities
state,number,rate,ageAdjustedRate
Alabama,342,7.2,6.9
Arizona,489,7.7,7.4
Arkansas,206,7.1,6.4
California,2238,6.0,6.5
Colorado,305,6.1,7.2
Connecticut,237,6.6,5.3
Delaware,57,6.3,5.8
District of Columbia,26,4.3,4.6
Florida,1755,9.3,6.7
Georgia,472,4.9,6.3
Idaho,123,7.8,8.3
Illinois,909,7.1,6.9
Indiana,485,7.5,7.2
Iowa,320,10.5,8.1
Kansas,256,9.0,7.9
Kentucky,296,6.8,6.8
Louisiana,265,5.8,6.2
Maine,137,10.3,8.2
Maryland,389,6.7,6.9
Massachusetts,459,7.0,5.9
Michigan,816,8.3,7.4
Minnesota,512,9.7,8.9
Mississippi,174,5.9,6.1
Missouri,477,8.0,7.2
Montana,94,9.5,8.1
Nebraska,183,10.0,8.7
Nevada,149,5.5,6.6
New Hampshire,116,8.8,8.0
New Jersey,645,7.3,6.5
New Mexico,171,8.3,8.3
New York,972,5.0,4.5
North Carolina,636,6.7,6.8
North Dakota,61,9.1,7.0
Ohio,920,8.0,6.9
Oklahoma,249,6.6,6.4
Oregon,356,9.3,8.3
Pennsylvania,1184,9.3,7.1
Rhode Island,96,9.1,7.1
South Carolina,381,8.2,8.3
South Dakota,86,10.6,8.3
Tennessee,435,6.9,6.9
Texas,1492,5.9,7.6
Utah,170,6.2,8.9
Vermont,70,11.2,9.6
Virginia,520,6.5,7.0
Washington,514,7.6,7.9
West Virginia,142,7.7,6.2
Wisconsin,492,8.7,7.4
Wyoming,37,6.6,6.8