A recent article I posted about abnormal lab test results showed how they can sometimes lead to a diagnosis, but more often than not take you down a primrose path to nowhere. My premise was that slightly abnormal tests should not be ignored. They should be repeated to confirm if they are really abnormal. If they are, then a work up to determine why is in order.
Another approach to the abnormal lab result dilemma says that there are certain tests for specific problems that may indicate real disease even if they are only a little abnormal. These tests are highly sensitive meaning when they are abnormal, they can say the patient has a problem even if the test is only abnormal by a few points. An essay on medscape.com addresses this problem and cites four examples, but I’m limiting discussion to three. They are:
TSH (thyroid stimulating hormone) levels for hypothyroidism
HbA1c for prediabetes
Testosterone levels for hypogonadism
Most normal “ranges for lab tests are set by taking the middle 95% of values from healthy individuals.” This is the classic “bell-shaped curve” where you have the results of a specific test for 1000 patients. Most of the results fall into the middle of the range of results. For example, for one test, healthy patients have a range of results from 1 to 100 with most falling in the middle of the range between 25 and 75. To set the “normal range” the lowest 2.5% and the highest 2.5% are eliminated leaving the normal range between 3 and 97. The 3 to 97 normal range is the middle 95% of the results. So just because your result falls outside of the normal range, it doesn’t men you have the problem that test may indicate.
A slightly high TSH could indicate you have an under active thyroid. But you have no symptoms of hypothyroidism and your T3 and T4 thyroid hormone levels are normal. So you wait a month and repeat the TSH level to see if it’s still abnormal. This time it is normal. You are not hypothyroid.
The same is true for A1c and testosterone levels. A very slightly elevated HbA1c doesn’t mean you have prediabetes, and a slightly low testosterone doesn’t mean you have hypogonadism. Other factors such as “symptoms and history” must be considered and included in the decision to make a diagnosis.
[An aside: Manufacturers of testosterone replacement products hyped the possibility of natural under production of the hormone to sell their testosterone replacement products. Hypogonadism was way overblown. Millions of tired, aging men had normal, or borderline low, testosterone levels and were mistakenly called hypogonadal. “Treatment” with testosterone replacement products failed, in my experience, to make these so-called hypogonadal men better.]The normal range for lab results is arbitrary. It’s the middle 95% of all results from healthy patients. But the 2.5% above and below that normal range may be just as normal as someone right in mid range. Those patients with results farther outside of the normal range will begin to have symptoms and indications of the presence of the disorder. Patients are sometimes told they have a disease they may not really have. This can affect insurability so it’s important to be certain of a lab result before committing to a diagnosis.
Reference: Dadhwal US. The Medicalization of Normal Lab Variations: When Abnormal Isn’t Disease. medscape.Com 2026 April 28.
