Smoke and drink coffee to avoid Parkinson's --or maybe not

From MedPage Today:

In a case-control study of those with Parkinson's in families, those with the disease were less likely to smoke or consume large amounts of caffeine than other family members, reported William K. Scott, Ph.D., of Duke, now at the University of Miami, and colleagues, in the April issue of Archives of Neurology.

But as a couple of people point out in the comments section, this finding doesn't prove much:

I wonder if they considered the possibility that people who are likely to develop Parkinson's disease may react differently to the pleasures of coffee and cigarettes?

and

Did it ever occur to these geniuses that people who develop Parkinson in latter life didn't drink coffee when they were younger because they had more of a tendency to become jittery than people who don't eventually get Parkinson's

and

If we can connect the dots, why are we not seeing the dots connected in any of the scientific publications?

This is beyond my competence but seems like an interesting line of discussion.