correlation and causation

"CASA found that 90% of adults who have addictive disorders started using substances before the age of 18. In comparison, only 4% of Americans who abuse as adults started using these drugs when 21 or older."

This seems like a classic example of correlation ≠ causation (or at least not necessarily). The type of people who don't use drugs until 21, are also the type of people who have good self-control. If the addicts waited until 21, they might still have equally poor self-control, and still get addicted later in life.

Very true, i waited until after i was 18 to smoke/drink and it is not somthing i have to have or somthing i abuse. I have a couple drinks and stop, Its easy.

I agree that there are problems with the study, but none of the sources I found commented on the methodological problems. And it was not my place, as I understand wikinews, to point out problems that I saw but that are not found in the sources.