We've already mentioned scattering (by yeast and trub) especially in intruments which measure actual refraction (as opposed to reflection in a digital instrument). Anything in the solution which isn't sucrose (which is what they are calibrated for). In #58 at
I've listed refractive index for sucrose, glucose, fructose and inverse sugar. As those data show the change is small but other dissolved things like salts, hops oils... would also have an effect. Can't think of anything else at the moment.