The press release in your link is actually a correction. The Center based a recent study on a flawed household sample that resulted in the estimate for underaged drinkers being doubled from 11% to 25%. The latter was the figure hyped by the media nationwide. Of course, the correction received almost no coverage.