A note on a result of Rousseau for concentration measures