Reference ranges: I don't understand why the... - Thyroid UK

Thyroid UK

111,422 members129,552 posts

Reference ranges


I don't understand why the reference range for the particular laboratory analysing the blood is needed in interpreting a result for TSH/T4, etc.

e.g. "There is a range, which is used for the tests. These vary sometimes which is why you must always ask what the range is, so that you can see where you are in the range."

A range of what, exactly?

I thought a blood test result, e.g. TSH of 2, was an exact measurement of hormone level in that particular blood sample, and the reference range was just what the lab recommended a person's results should fall into.

If one needs the reference range, does that mean instead that a person's result is predicated as falling into a particular percentile of the population as calculated by the laboratory? In other words, if exactly the same blood sample was sent to ten different laboratories, you could get up to ten different results back?

Like, say, if you set an examination with 100 participants. You could either:

☻Give each student an exact mark, as in "Your score was 75%", or

☻You would look at the lowest mark scored and the highest mark scored and grade accordingly, so if the highest mark scored was 85% and the lowest was, say, 35%, you'd award scores between 75% and 85% Grade A, and those scoring 35% to 45% would be in the lowest grade, even though no-one scored zero.

How much does the "range" matter, particularly given they are so wide? I mean, if someone said to me, "My TSH is 1", I'd respond: "That sounds about okay, could perhaps be a tad lower: how do you feel?" But if they said "My FT4 is 1", I'd go: "Eek!".

I hope what I'm asking is clear. I've spent some time browsing through posts on reference ranges and ended up more confused than ever. Can someone please explain in very simple layman's terms? Many thanks.

18 Replies

Yes, thank you, I understand all that.

But what I'm asking is whether a blood test result is an exact measurement of hormone level in that particular blood sample, or an indication of where it falls in the range of that particular laboratory?


It would be much easier if ranges were identical but Laboratories have different machines with different ranges. So it could mean the difference between being diagnosed or not.

I cannot give you an explanation but doctors are told to diagnose by the TSH alone and adjust the medication (if you are on any) to keep the TSH 'in range' even though it can vary throughout the day that's why we recommend a blood test as early as pos as TSH is highest. So I don't see that as a 'perfect' test as many patient's dose is up or down constantly due to their TSH level with the result they canot get stabilized. If I get a TSH blood test at 9 a.m. and it is 'out' I might be diagnosed. If I have a blood test at 2 p.m. I may not even though I have clear clinical symptoms. Some patients' TSH doesn't rise sufficiently to be diagnosed at all.

This is what Dr Toft of the British Thyroid Association states:-

"The aim of thyroxine replacement therapy is to normalise plasma TSH and to achieve a clinically euthyroid state. To obtain this, FT4 and TT4 have to be maintained at, or just above, the upper reference interval".

NB: Dr. Toft, Consultant Physician, states in the British Thyroid Foundation newsletter Issue No. 23 that normal ranges are: T4 (10 – 25) and TSH (0.15 – 3.5). He also states that "the correct dose is that which restores good health; in most patients this will be associated with a level of T4 in the blood towards the upper part of the normal range or even slightly high and a TSH level in the blood which is in the lower part of the normal range."

First very rarely is FT4 and TT4 tested to know whether they are maintained at the upper (or even FT3 which is the most important) and it's not even mentioned above.

Second, if we are kept in the above 'normal range of 0-15 - 3.5 the 3.5 might be too high for us and the doctor won't increase medication because we are in 'range'. The BTA guidelines insist that our TSH must reach 10 before being diagnosed (a conflict if there ever was one).

The 'old fashioned' method of diagnosing was without blood tests, taking the history and clinical symptoms of the patient into account, plus their 'look' and prescribing Natural Dessicated Thyroid hormones. The country would save hundreds of pounds if they used this method because people would be diagnosed more quickly, be given a trial of medication according to symptoms to see if there was an improvement: prescribed medication until they felt well (not until they were 'in range' and meds not increased).

SailingBy in reply to shaws

Thank you very much. I wish I could find a way of phrasing my question so that what I mean is a little clearer. If we could leave to one side the rights and wrongs of TSH diagnosis and the benefits of testing for T3/FT4, etc, and how they ought to do it, and look at the actual result -- not necessarily TSH, but anything. If my blood sample was sent to ten different laboratories, could you conceivably get up to ten different results back of the actual measurement, the figure, of whatever they are testing for, leaving aside whatever "range" they set as ideal?

diogenes in reply to SailingBy

The answer to your question is probably "ten different ranges". This is because the mix of healthy subjects/patients the individual lab uses to establish its range depends on the exact mix, and the diagnostic peculiarities of the particular doctors supplying the diagnosed patients. The ranges won't be terrifically different, but the upper and lower limits may alter. Your actual result could vary by up to 25% (the present intrinsic variability in TSH tests from the manufacturers).

SailingBy in reply to diogenes

That's brilliant, thank you very much, most helpful.


Hopefully someone who is au fait with blood levels will respond later. I know what you mean, i.e. if I had five blood tests in five different labs, would my number be the same from each lab omitting the ranges?

SailingBy in reply to shaws

Yes, that's it exactly, thank you. It's whether any laboratory's range is just their scientific opinion, almost, of where a blood test result should lie, depending on what demographic they use to analyse it.

PR4NOW in reply to SailingBy

If you sent a single sample to 20 different labs you would probably get 20 different results with 20 different ranges. This has already been tested and established. The bugger is, what does the result really mean? The answer is, for a diagnostic purpose not much unless it is at the extremes. PR


TSH ranges tend to be similar - though we do see some variation at both ends of the range.

FT4 varies hugely. At one end you see something like 7 to 13, and at the other end, 12 to 23. (I am not 100% sure whether or not there is a gap between the top of the lowest range and the bottom of the highest range - or a very small overlap.) Someone saying "12.5" could be just above the bottom of range or just below top of range. Further, you example of FT4 = making you go "Eek!". Why would it do that? A USA range could well be 0.8 to 1.6! (They use different units.) And we do see a steady stream of USA results posted here. Unless you know the range in use, any interpretation is based on the assumption that the range is similar to what you are familiar with.

T3 represents another opportunity for confusion. A Total T3 range is typically 0.9 to 2.6 nmol/L (USA uses ng/L which allows for further confusion.) A Free T3 range is typically 3.3 to 7.1 pmol/L. We have had several occasions on which someone has posted a T3 result and caused confusion. Is 2.9 high or low? Actually, first we need to ask, is that a Free or a Total T3 test?

Complex medical tests are miles away from the phenomenal accuracy we have come to expect of physics and chemistry. Tests can be compromised by their technology, their susceptibility to interference, the actual conditions. Diogenes has explained that FT4 testing must be done at blood temperature. This has not always been fully recognised.

Some years ago it was appreciated that high TSH can be caused by macro-TSH (an antibody attached to a molecule of TSH). Some TSH tests were modified to exclude that affecting the result. (I don't know for sure if any UK tests are still affected by this - I think so.) If someone has discordant TSH results, one of the first things a testing expert could is to get samples taken at one time analysed in different labs which use different technologies. The differences they can sometimes find are little short of astonishing.

We end up saying:

Tests vary.

Ranges vary.

Units vary (though not so often within the UK).

Reference ranges (preferably including measurement units) can help to avoid mistakes as to which test is being discussed.

This is a very much less than satisfactory state of affairs.


SailingBy in reply to helvella

The point I was—clearly unsuccessfully— trying to make was that, all things being equal, we know, broadly speaking, whether someone's TSH or FT4 is way off the scale, and my query centres around why therefore we need to know the range given by that particular laboratory. For the avoidance of doubt, I meant if I was talking to someone in the UK with a test done in the UK. I don't doubt that confusion may arise between US and UK measurements or whether someone is talking about free or total T3, or that results can vary.

I'm running out of ways of how to word my question. Is the laboratory's range based on the machine they do the testing on? Or is it a guideline based on a survey of actual people, however this may be calculated, and however accurate or flawed?

helvellaAdministrator in reply to SailingBy

I think I have failed to make it clear that the FT4 ranges I referred to are WITHIN THE UK.

That is, what do you make of an FT4 result of 12? Is it high or low?

The ranges are based on a combination of the technology, the maker of the analysis equipment and kits, and the lab's interpretation of the results they get - possibly including testing real people.

Same blood but different labs = probably different results.


humanbean in reply to helvella

Re: your comment "Same blood but different labs = probably different results." I posted something below that is relevant to this. Thought you might find it of interest.

I don't know whether this will be helpful to you or not.

Last year I had a Blue Horizon Comprehensive test done and a Genova Diagnostics Thyroid Plus test done on the same day. The needle was stuck in my arm once, and all the tubes needing to be filled for both sets of tests was filled from the one needle.

Two of the tests were common to both Blue Horizon and Genova Diagnostics - TSH and Free T4. So I got them tested by two different companies from a sample that was (as near as realistically possible) the same sample.

The results :

Genova Diagnostics

TSH - Reference range 0.4 - 4.0 mu/L My result = 3.8 This result is 94% of the way through the ref range.

Free T4 - Reference range 10 - 22 pmol/L My result = 12.5 This result is 21% of the way through the ref range.

Blue Horizon

TSH - Reference range 0.27 - 4.2 mu/L My result = 4 This result is 95% of the way through the ref range.

Free T4 - Reference range 12 - 22 pmol/L My result = 14.21 This result is 22% of the way through the ref range.

The two TSH ranges were different, and so were the Free T4 ranges. My results were different from the two sets of tests. But when looking at how far of the way through the reference range my results were as a percentage they agreed very well - 94% and 21% from Genova, 95% and 22% for Blue Horizon.

I actually found this encouraging. I always use the "percentage through the reference range" as a way to compare tests from one occasion to the next - it helps to even out any changes in reference range.

helvellaAdministrator in reply to humanbean

Thanks, humanbean. I agree that percentage-through-range should be reasonably consistent, as you have shown. The problem is that we have relatively little evidence for that as not so very many people have tests fro different labs which they think should reflect each other.


SailingBy in reply to humanbean

Very interesting, humanbean, thanks for posting!

HI All Labs have different ranges so interpret results you need the range for the Lab where blood analyzed.


I should like any lab to give an answer in percentages. It would be easier then to see where FT3 is in relation to FT4. Always mean to work it out but never do!

I should like any lab to give an answer in percentages. It would be easier then to see where FT3 is in relation to FT4. Always mean to work it out but never do!

You may also like...