Before you poo-poo this as blindingly obvious, research is just that, a study based on evidence. It's all very well laughing at the no brainier hypothesis, or hypotheses, but research rules out coincidence in a search for correlation and cause. Otherwise we have a hypothesis that cows eat only green things, based solely on the coincidence that grass is green.
As an example, the study may have shown that cyclists are fitter and slimmer because they are of higher average intelligence and therefore take more care of themselves, and the cycling is a by-product of that lifestyle mchoice, not the cause. Or you could make up and insert your own semi-spurious coincidence at this point.
What's more, research around health statistics may have more important downstream ramifications. For example, if it is proven beyond doubt that cyclists in general are fitter and slimmer, then someone may look at the savings in healthcare per cyclist, then how to manage or maximise that saving by increasing cycling infrastructure or (God help us) driver education. If the correlation is only weak or coincidental, then it's harder to argue that encouraging more people to cycle would be beneficial, and if you are commissioning a multi-million pound highways budget then you're going to need more than "bleedin'obvious innit" as your mainstay argument.