Screening for NIDDM: increasing evidence that it is necessary Part 1: Should it be done?