For simplicity, lets use our Sun as the star.
Next, a dyson sphere is built around it (completely enclosed solid sphere, not a dyson swarm or network, etc.). Handwave how it's built, it's not relevant to the question.
Not sure if the diameter of the dyson sphere makes a difference, but in case it does, lets say its inner surface is about 1AU from the Sun. (sorry Earth, you've been replaced).
Again, for simplicity, lets assume no other celestial bodies remain inside the sphere (Mercury, Venus, other small bodies, etc., completely used up to build the sphere, or otherwise intentionally removed)
Now, the supposed purpose of a dyson sphere is to collect the energy from the star, and put it to good use doing ... something. But shortly after completion of this one, its builders (for whatever reason) flipped the switch from "collect" to "reflect". So instead of collecting the energy from the Sun, it's reflecting most of it back.
What about efficiency? Well, silver and can be up to 95% reflective, at least across the visible spectrum. Aluminum comes close at about 90%. I don't have numbers for materials reflecting non-visible spectrum energy, but since we're talking about dyson sphere level technology, lets handwave the composition of the reflective surface, and just say it reflects 90% of all forms of energy that strike the surface. Regardless of what form of energy it is, the surface reflects 90% of it back at the sun.
The remaining 10% of the energy, the part not reflected, is absorbed (collected) by the sphere, converted to heat, and then radiated into space on the outside of the Dyson sphere, and has no ill effect on the sphere itself.
Does this cause any changes in the star?
Best answers will include details of changes, if any, and explanations of why they take place, and the time-scales during which the changes manifest themselves. Or, explanations of why no changes take place.