Occurrence and management of an aberrant free T4 in combination with a normal TSH

[helvella]

One of the more common questions we see is why Free T4 levels do not always correspond with TSH. Although in many ways this paper is not that important, it does provide a neat, simple check list of possible reasons for FT4 being higher or lower than would be predicted by TSH alone.

The very fact that it says this is even possible further undermines the use of TSH-only. After all, we are told repeatedly that TSH is effectively perfect. Then this paper says FT4 could be high or low despite the apparently good TSH result.

We also need the parallel study into high or low TSH results with good FT4 levels!

The authors do appear to be aware - see Limitations in the full paper.

Neth J Med. 2018 Sep;76(7):314-321.

Occurrence and management of an aberrant free T4 in combination with a normal TSH.

van Veggel KM1, Rondeel JM, Anten S.

Author information

Abstract

BACKGROUND:

Thyroid function tests may show the combination of a normal concentration of serum thyroid stimulating hormone (TSH) and an increased or decreased level of free thyroxine (free T4). How often this occurs is unclear and not everyone is familiar with how it should be adressed.

METHODS:

We conducted a retrospective cohort study of all adult patients who presented at a non-academic general hospital in the Netherlands between 1 January 2010 and 31 December 2014 and yielded an increased or decreased free T4 in combination with a normal TSH. Exclusion criteria included the use of thyroid medication, pregnancy, a history of thyroid surgery and treatment with radioactive iodine. The medical records of the patients included were retrieved and evaluated.

RESULTS:

Of the 30,143 combined TSH and free T4 measurements in 23,199 individual patients, 1005 measurements (3.33%) in 775 patients (3.34%) yielded an aberrant free T4 in combination with a normal TSH. 398 patients (1.72%) had a persistent aberrant free T4, 349 (87.7%) of whom had a decreased free T4 and 49 (12.3%) an increased free T4. In 58 of the 398 patients (14.6%) with a persistent abberant free T4 a possible cause was established by the treating physician. However, upon re-examination of medical files a possible causative factor could be identified in 123 patients (30.9%).

CONCLUSION:

In our study population the prevalence of hyperthyroxinemia or hypothyroxinemia in combination with a normal TSH was 334 per 10.000 patients. When records were thoroughly searched, identification of potential causative factors increased substantially. Clinicians should be encouraged to check for underlying causes.

PMID: 30220656

ncbi.nlm.nih.gov/pubmed/302...

Full paper freely accessible here:

njmonline.nl/getpdf.php?id=...

[Aurealis]

Thank you for posting. What an interesting article, I wish it had been about 25 years ago. But surely useful now to those experiencing difficulty obtaining treatment / adequate treatment. I wonder how many GPs will read this?

[helvella in reply to Aurealis]

How many fingers do you have?

[Hidden in reply to helvella]

LOL, I had to read this twice to get it. I'm clearly under-caffeinated.

[lady_eve in reply to helvella]

On one hand?

[jimh111]

Thanks for posting. I've only superficially skimmed it. It's a disappointing study.

If TSH and fT4 were independent variables we would expect 5% of fT4 to lie outside the reference interval, i.e. be classed as 'abberant'. They found a smaller percentage.

Of course TSH, fT3 and fT4 are all interrelated and vitally patients can be hypo or hyper when one, two or all three of these parameters are within their reference intervals. e.g. TSH=1, fT3=3.5, fT4=12 and the patient is severely hypothyroid.

So, although this paper gives a few examples of rare causes of thyroid disorders it reinforces the attitude that the the three parameters are independent and that simplistic interpretation of blood tests is infallible.

I'm not shooting the messenger! This gives us an opportunity to show how seemingly helpful studies are reinforcing bad science. Interpretation of thyroid function tests is very difficult. The studies carried out by the Thyroid UK advisors demonstrate how patients can suffer from thyroid disease whilst having seemingly innocent TFTs.

[helvella in reply to jimh111]

There is a lot wrong with the paper.

Start with the concept of "normal" TSH just being TSH within range. Continue with somehow missing biotin interference with tests. Indeed, the paper gives the impression that there is no such thing as assay interference - just a real impact on FT4. And although they do actually mention FT3, the almost total absence of results meant that there was almost no thought about it.

Again and again and again, we see TSH-only as impairing not just the individual patients who are not properly investigated but also research - even at this relatively straightforward level. With even worse effects when we consider meta-analysis.

Despite all this, the list of identifiable causes, neatly presented, could be of some benefit.

[Aurealis in reply to helvella]

There’s just the worry that the term ‘acute psychosis’ could be applied a bit too liberally to those (quite rightly but inconveniently) ‘pushing’ for treatment, for T3 etc.

[Hidden in reply to jimh111]

I am coming to the conclusion that a lot of medical papers are weak. In themselves, they're fine - they give some information about something specific - the main issue as I see it is that more importance is put on that information than it warrants. (It isn't entirely uncommon to see data collected that says A, and a conclusion that says B.)

The main issue, as I see it, is that this simply does not apply to us - we diagnosed thyroid patients taking replacement hormones. "Patients were excluded if they used thyroid hormone replacement (for example levothyroxine and liothyronine) or thyroid inhibitory medication (for example propylthiouracil and thiamazole), had undergone thyroid surgery

or had been treated with radioactive iodine in the last 2 months" (p. 315).

[helvella in reply to Hidden]

A lot of papers are weak.

Nonetheless, in this case, I do suspect that the list of possible causes is reasonable. Maybe incomplete, for several reasons, but if you have a patient with discordant test results and one or more of the identified factors, that should be considered.

The paper does at least mention thyroid hormone resistance (impaired sensitivity to thyroid hormone).

[jimh111 in reply to Hidden]

"RAI during last two months" So, half a thyroid will give the same TFT results as a full one. Why do we bother to grow both lobes of the thyroid?

[shaws]

Thanks for posting helvella. It is about time Endocrinologists read research articles instead of permitting constant suffering by many patients who're given other medications for the symptoms.

"The very fact that it says this is even possible further undermines the use of TSH-only. After all, we are told repeatedly that TSH is effectively perfect. Then this paper says FT4 could be high or low despite the apparently good TSH result."

[lady_eve]

I'm pleased to see central hypothyroidism included in Table 1. It only took 14 years for a consultant to spot mine, based on my TSH vs. T4 results from the beginning!

[humanbean]

I'm probably being dim here, but Table 1 seems to be written very poorly.

In answer to Column 1, Row 1 :

Causes of a normal TSH in combination with a decreased free T4

two of the answers are "estrogen and "growth hormone". Everyone has some estrogen and some growth hormone. So what do they mean? High levels or low levels lead to normal TSH and low T4?

[helvella in reply to humanbean]

I agree that it is not as clear as it might be but, as other substances include heroin, I think it has to be read as "higher than usual levels".

[humanbean in reply to helvella]

Doh! I never thought of that! Thanks.