It's not impossible,
This is not well studied. Because if they're saying the "fever" is the correlation, they don't really know if it's just "infection" which does it, or the "fever."
So, they would need to control between women who had "infections" and with/without taking fever reducers.