"Espay and his colleagues theorize that Parkinson's is not one disease but rather several diseases when considered from genetic and molecular perspectives." We've been saying this on this forum for years.

"Espay and his colleagues say the field must work to develop an ideal set of biomarkers." Biomarkers, YES! But an ideal set of biomarkers will take years to develop, actually delaying the progress which it is intended to accelerate. We should use the best thing that we have now: automated, objective tests, e.g. tap tests.

I would focus on personalized, n=1 clinical trials. This takes into account the fact that what I'm interested in is whether a therapy works for me. The fact that it doesn't appear to work for you is not, in itself, important for me. Historically we have needed to use large population sizes in order to get statistical significance. But we can get around this requirement if we make high frequency observations. For instance, rather than test a subject every 3 months, we could test a person, a thousand times per day.

John