Thyroid Blood Tests Part 1 - Reference Ranges

A laboratory reference range is the range in the unit of measurement for the given test that is usually provided with a specific test result. It is usually expressed as lowest level - highest level. Thyroid patients would be well advised to insist on having the reference range given to them for each and every test result they are given. Keeping a record of all test results including the reference ranges for each is very useful. A reference range may also be called a reference interval. Laboratories assess a wide population of people when calculating the reference ranges for TSH, FT4 and FT3. They try to avoid including those people with known thyroid disease, although some people may inadvertently be included that have undiagnosed health issues, including mild thyroid abnormalities.

The laboratory ranges for TSH, FT4 and FT3 thyroid hormones are defined statistically to use a 95% interval such that approximately 95% of the population will fall within the reference range. It makes sense for the laboratories to do it this way because thyroid hormones do not have a Gaussian (normal) distribution and a 95% interval gives a clear, unambiguous number. Now, inevitably, some people will not fall within these reference ranges simply due to the statistical methods used to create them in the first place. These people will have normal levels of thyroid hormones for them but these levels will not conform to the laboratory reference ranges. 

Statisticians studying a large number of samples of a population, sometimes identify samples that they refer to as outliers. Outliers are results that are on the extremes of a population. Outliers are either due to measurement errors or due to the extended tails of a distribution. Many of these outliers are therefore real results that are perfectly normal results for an individual. These results will not be included when the lab reference ranges for TSH, FT4 and FT3 are set.

In addition, because there are practical limits on how many people can be assessed in the population to form these ranges, there inevitably will be people who have not been sampled as part of this process who would be significantly outside the typical range of results, i.e. if they had been sampled they would have been likely to have been classified as outliers.

So, firstly, there inevitably will be some people who may have perfectly healthy TSH, FT4 or FT3 levels that happen to fall outside of the laboratory reference ranges. Secondly, the population used to create the reference ranges does not contain only healthy people. Therefore, it is likely that the reference range is further restricted because it is not derived from entirely healthy people.

Both of these conclusions may mean that if an unlucky patient happens to be one of those people who requires a level of one of the thyroid hormones, at, or beyond, the normal range in order to feel healthy, then during thyroid hormone treatment this patient might reasonably expect to continue to feel unwell until this healthy level was duplicated once more. Simply, telling the patient that their hormones fall somewhere within the reference range is no guarantee that they are adequately treated.

To complicate matters further there is an important mathematical relationship between TSH and FT4 and FT3. TSH exponentially increases as FT3 and FT4 levels fall. This isn't a simple linear relationship at all. It is based on something called and "an inverse log - linear relationship between TSH and FT4".  In simple terms, if a doctor were to point out that at a TSH of, say, 2.0 you don't need a medication increase, then this may not actually make sense. Because of the actual mathematical relationship between TSH and FT4 it is likely to take the same amount of thyroxine to reduce the TSH from 2.0 to 1.0 as it does to reduce TSH from 10.0 to 5.0, i.e. it may still require a huge amount of thyroid hormone to reduce TSH to 1.0 and the patient's doctor may have more 'room for maneuver than they suspect.

TSH distribution is also skewed to the left with the median TSH being about 1.0. Similarly the FT3 and FT4 distributions are also skewed with most people being above the halfway point. These points need to be remembered when patients still don't feel well but there laboratory test results look 'OK'. There may be far more room for adjusting the patients dosage even for those doctors that rigidly adhere within the laboratory reference ranges instead of many other potentially more useful measures. I credit Jim Harwood for reminding me about the mathematical relationships just mentioned.

The application of an individual's thyroid blood test results, which concludes that the person's thyroid hormones are normal because they fall somewhere within the reference ranges is unlikely to be applicable to 100% of patients on T4 or T4/T3 because of the statistical methods used to create the ranges in the first place.  Some patients that need additional or different treatment will not be served by rigidly adhering to treatment that remains within the reference ranges.

This laboratory-test approach that modern medicine uses is more suited to a fast, conveyor-belt, production line processing of patients than it is to providing each, individual patient with advice and help that is uniquely tailored to them.

All of the above is relevant but does not even account for those people who fail to become well through the use of T4 or T4/T3 (or even on T4/T3 require a large amount of T3).

Once a patient is on T3 only treatment the above issues are only a subset of the potential issues that they may face with thyroid laboratory tests.

Since thyroid blood tests are such an important topic to thyroid patients I will be writing at least three more parts on this topic over the coming weeks.

Best wishes,

Paul