The correlation between two things doesn’t imply that one causes another. Knowing one variable may allow us to predict the other one, but it doesn’t tell us why there’s a connection.
For example let’s consider the following statement (regardless of whether it’s actually true):
Most programmers are male.
What this statement means is that a person who writes code is more likely to be a man than a woman. But this doesn’t mean that being a male causes people to become programmers (and definitely doesn’t mean that being a programmer causes people to be male).
It tells us what to expect, but it doesn’t reveal the mechanism that influences these factors. In the previous example, the correlation may be high, but the relationship is not causal.
What may be happening under the hood is the presence of a third critical factor.
Consider the situation where the people who regularly take the vitamin X have less heart problems than the people who don’t. That can lead you to believe that the supplement has a positive effect on the heart’s health, but there is a flaw in that logic.
What if the vitamin X costs a lot of money, and its consumers are mostly wealthy people, who generally tend to eat healthier?
The numbers don’t lie. But be very careful when you interpret them.