A remark on the minimal index of subfactors