This is an interesting question.

I'm not qualified to answer the specific question of the thread's starter, since the answer must, at least in part, be contingent on the general health of both the father and the child. Instead, let's look at the general case.

A paper by Payami et. al. gives the relative risk factors for parents of early onset PwP as 7.76 and for parents of late onset people as 2.95.

Note they give data for the parent given the child having PD, whereas we are really interested in the opposite, the child given the parent having PD. We can use Bayes Theorem to move from one to the other. If we assume that prevalence rates are not changing, we get the same values either way: that is, a child with a PwP parent has on average between 3 and 8 times the risk of having Parkinson's than that of someone who doesn't. (Note this paper is 10 years old, advances in genetics most likely mean that estimates can be more precise.)

However, the news is not as bad as this may, at first sight, suggest. PD is not a common disease, so these factors only take the probability of getting PD from the very unlikely to the unlikely, rather than the likely. A graph in the same paper gives the cumulative probability of escaping PD as 98% (no PD child), 94% (late onset child) and 86% (early onset child). And, those figures are for people who survive into their 80s; many people will die earlier than this, often before being diagnosed.

Reference

"Familial Aggregation of Parkinson DiseaseA Comparative Study of Early-Onset and Late-Onset Disease"
Haydeh Payami, PhD; Sepideh Zareparsi, PhD; Dora James, BS; John Nutt, MD
Archives of Neurology, May 2002, vol 59, no. 5
http://archneur.jamanetwork.com/arti...ticleid=782004

John