But what about the highest 25 and lowest 25 that were originally a part of group of 1,000. By definition, you've just made them abnormal, even though the group was supposedly all normal before plotting out the results.
Do this over & over again for each test result and start increasing the total number of subjects from 1,000 to 10,000 to 100,000, etc. And don't forget to throw out the highest 2.5% and lowest 2.5% each time so that you're always left with the middle 95% that then forms the normal bell shaped curve.
I give you this long winded prelude as background for a study to be published next month in JCEM in which 10,083 women and 5,023 men free of baseline thyroid disease were followed for 11yrs after noting that their baseline TSH (thyroid stimulating hormone) was normal and within the normal reference range.
And yet, after over a decade of follow up, even though they all had normal TSH initially, those with normal TSH closer to the upper limit of normal were at greater risk for developing hypothyroidism. And those with normal TSH closer to the lower limit of normal were at greater risk for developing hyperthyroidism. Which goes to show you that Goldilocks was onto something. You want your TSH just right, not too high and not too low, even though it's normal.