You will have heard the expression before, "correlation is not causation" in your biostatistics class. In epidemiology, since we measure the association between exposures and outcomes, we often say that association does not mean causation, that is, because an exposure might be associated with a health outcome, it doesn't mean that the exposure caused the outcome. This basic assumption is why the specific cause of a disease is not always easy to determine, especially for chronic, non-infectious diseases. Imagine you see a correlation between the number of PhDs in computer science awarded every year and the income from video game sales. As the number of PhDs awarded goes up, so does the income from video game sales.