Hydrogels are sponge-like materials that can absorb or expel significant amounts of water. Swelling up from a dried state, they can swell up more than a hundredfold in volume, with the kinetics and the degree of swelling depending sensitively on the physicochemical properties of both the polymer network and the aqueous solvent. In particular, the presence of dissolved macromolecules in the background liquid can have a significant effect, as the macromolecules can exert an additional external osmotic pressure on the hydrogel material, thereby reducing the degree of swelling. In this paper, we have submerged dry hydrogel particles in polymer solutions containing large and small macromolecules. Interestingly, for swelling in the presence of large macromolecules we observe a concentration-dependent overshoot behavior, where the particle volume first continuously increases toward a maximum, after which it decreases again, reaching a lower, equilibrium value. In the presence of smaller macromolecules we do not observe this intriguing overshoot behavior, but instead observe a rapid growth followed by a slowed-down growth. To account for the observed overshoot behavior, we realize that the macromolecules entering the hydrogel network not only lead to a reduction of the osmotic pressure difference, but their presence within the network also affects the swelling behavior through a modification of the solvent-polymer interactions. In this physical picture of the swelling process, the net amount of volume change should thus depend on the magnitudes of both the reduction in osmotic pressure and the change in effective solvent quality associated with the macromolecules entering the pores of the hydrogel network. We develop a phenomenological model that incorporates both of these effects. Using this model we are able to account for both the swelling and compression kinetics of hydrogels within aqueous polymer solutions, as a function of the size of the dissolved macromolecules and of their effect on the effective solvent quality.