Do epidemiological data on exercise and substance abuse support prescriptions for prevention?

A Twitt from the Foundation for Alcohol Research (@AlcoholResearch) today struck my attention. It sounded to me like the usual slippery slope of creating human health prescriptions from limited scientific findings.

Teen athletes may drink more, but smoke less & use fewer drugs. How do you lessen your teen’s risk? http://bit.ly/mysWna

The link is to their newsletter which overviews a paper by Terry-McElrath and O’Malley, currently in pre-print at Addiction. The overview is pretty straightforward, based closely on the paper and eschews the problem with the Twitt, which was the question as to whether you could “lessen your teen’s risk”. So I am mostly mollified.
The paper in question reports data from a survey of over 11,000 US high school seniors (classes of 1986-2001), captured as seniors and then followed longitudinally until age 26. These data were collected as part of the Monitoring the Future study which we discuss quite frequently on this blog.
The key focus of this paper is on the amount of physical activity the surveyed HS seniors reported at first contact. The Participation in Sports, Athletics or Exercising (PSAE) measure was derived:

…by asking, “How often do you actively participate in sports, athletics or exercising” (1=never, 2=a few times a year, 3=once or twice a month, 4=at least once a week, 5=almost every day).