Statistically significant just means that the results are correlated at greater than some probability, not that the effect itself is significant.

I.E. you can have a statistically significant but unimportant effect.