The interpretation of Thyroid Function Tests
Background:
The Health Study of Nord-Trøndelag (HUNT).
' European Journal of Endocrinology 2000 143 639-647. Download here,
or from
Norsk Epidemiologi 203; 13 (1): 19-32 Download here.
‘Normal’ TSH
By Alun Stevens MSc FIAA, Secretary of Thyroid Australia
The most common questions that Thyroid Australia is asked involve the interpretation of Thyroid Function Tests (TFT’s). Many people have been told that their TFT results are ‘normal’. So what is ‘normal’? In this article we will focus on the test for Thyroid Stimulating Hormone (TSH) which is the most common test ordered.
The ‘normal’ Reference Range for the test is intended to represent the range of values which can be expected in the healthy population – ie those without any thyroid ailment. The Reference Range is found by taking a sample population of healthy individuals and determining their TSH levels. The lowest and highest 2.5% of readings are excluded so that the Reference Range covers 95% of the healthy population. There are a number of different tests for TSH with different levels of sensitivity. They each have their own Reference Range. The most common tests generally have lower limits to their Reference Ranges around 0.2 to 0.5 mIU/L and upper limits from 3.5 to 5.0 mIU/L.
A recent study in Norway provides a good example of the use of the TSH test in practice.1 The study involved 65,000 people. They were asked questions about their thyroid status and those with a history of thyroid illness were excluded. The blood samples were tested for Thyroid Peroxidase Antibodies (which are an indicator of likely thyroid illness) if they produced a TSH reading greater than 4. Samples with positive antibody results were also excluded. The survey, therefore, attempted to exclude people with any indication of thyroid illness, but still included those with Thyroid Peroxidase antibodies whose TSH reading was 4 or less. The TSH test kit used for the study had a nominated Reference Range of 0.2 to 4.5 mIU/L.
The results for women are shown in the chart. The results for men were only slightly different.
The features of this result are:
- The distribution of TSH readings in the healthy population is skew. It is not the common bell shaped curve centred in the middle of the reference range.
--- - The most common value, or Mode, is at 1.25.
--- - The Median value is at 1.50. This means that half the population (50%) have a TSH reading below 1.50.
--- - The average, or Mean, value is at 1.68. Over 60% of the population have a TSH reading below this value.
--- - The centre of the Reference Range for the test kit used in the study is 2.35. Almost 85% of the healthy population have a TSH reading below this value.
--- - The 2.5 percentile point (ie the point which excludes the bottom 2.5% of the population) is at 0.48. The 97.5 percentile point (ie the point which excludes the top 2.5% of the population) is at 3.6. The range between the 2.5 and 97.5 percentile points (0.48 to 3.6) is much narrower than the test kit’s Reference Range (0.2 TO 4.5).
--- - This narrowing of the range would suggest that the reference group used to calibrate the test kit possibly included people with some level of thyroid illness.
--- - This narrowing of the range between the 2.5 and 97.5 percentile points would potentially have been even more pronounced if all samples had been tested for Thyroid Peroxidase Antibodies.
The conclusions which can be drawn from this survey are:
- TSH results in the upper half of the Reference Range have a low probability of being ‘normal’. This does not mean that they are not ‘normal’. It means that they are unlikely to be ‘normal’.
--- - The Reference Ranges for TSH tests are potentially too wide, especially at the upper end. This suggests that ‘high normal’ TSH readings should possibly be treated with more suspicion than they currently appear to be.
--- - The centre of the Reference Range is clearly not a good target point because very few of the healthy population have TSH readings around this point.
--- - A much better target point would be around 1.0 to 1.5. But some people will feel better at higher levels or lower levels. This supports Prof Jim Stockigt’s view that the target should be a TSH reading around 1.0.2
Another important point which needs to be borne in mind when interpreting statistics like these is that it is the population which has a range of values with probabilities for each reading. Each healthy individual is only at one of the points. They are ‘normal’ when they are at that point. For those on thyroxine replacement, being in the Reference Range is not good enough in itself. You need to be at your own set point. This will probably be near the lower end of the Reference Range.
This analysis of the distribution of TSH readings in the healthy population supports our recommendations to thyroid patients:
- Obtain a photocopy of all your Thyroid Function Tests. Also get copies of the ones you have had done in the past. These copies will show both the readings and the Reference Ranges.
--- - When you are going for a new test, make a note of how you feel (especially make a note of any of the major symptoms of thyroid overactivity or underactivity), your weight and your dose. When you obtain your copy of the test result, write this information on the copy. Over time, this process will allow you to make an informed judgement in consultation with your doctor of what the correct set point is for you.
--- - Do not accept that a Thyroid Function Test is 'normal' just because the result is within the Reference Range if you are still feeling unwell.
References
- T Bjøro et al, 'Prevalence of thyroid disease, thyroid dysfunction and thyroid peroxidase antibodies in a large, unselected population. The Health Study of Nord-Trøndelag (HUNT).' European Journal of Endocrinology 2000 143 639-647. Download here.
- J Stockigt, 'Subclinical Hypothyroidism or Mild Thyroid Failure: How important is early diagnosis and what treatment is optimal?' Interview With Sigma Pharmaceutical 2001 http://www.thyroid.org.au/Information/Stockigt.html. Accessed 30 December 2001.
Alun Stevens is an actuary with his own consultancy. He is also Secretary of Thyroid Australia
This article is published in our newsletter Thyroid Flyer Volume 3 Number 1 January 2002.
This article can be reproduced provided it is reproduced in full, acknowledges the source and is not sold for profit.
(c) Copyright 2001, Thyroid Australia Limited ABN 71 094 832 023
PO Box 2575 Fitzroy Delivery Centre, Victoria 3065, Australia
Thyroid Australia Home www.thyroid.org.au 
