In statistics, which this is, a "prediction" is an outcome that the statistics suggests would happen. The "cause" isn't involved. A predicts B, means that having factor A means factor B is likely present also. And that is that. There is no suggestion of causation.

This is a science magazine, the reporter would have read a lot of papers over his career, and because "A predicts B" is such a common use to denote correlation in statistics, I wouldn't blame him/her.