yeah, it's not like this is a randomized experiment so there might be confounding factors. Observational studies aren't very useful for determining causation in general