A remark on the mean square distance between the solutions of fractional SDEs and Brownian SDEs