On urea's ability to stabilize the globule state of poly(N-isopropylacrylamide)