However, just because A is correlated with B doesn't mean A is caused by B. Maybe B causes A, maybe some other factor C causes them both, or it's remotely possible you've observed a coincidence.
Matt Ridley has a nice column for The Wall Street Journal on this very topic. His main example is that instead of increased carbon dioxide causing global warming, it now appears more likely that global warming is somehow implicated in an increased release of carbon dioxide.