Subclinical hypothyroidism

A study by Zheng 2022, in Frontiers (Ref. 1) looks at how you are able to characterize patients with subclinical hypothyroidism.

Comparison of Five Different Criteria for Diagnosis of Subclinical Hypothyroidism in a Large-Scale Chinese Population:

When an elevated TSH concentration was defined as ≥4.50 mIU/ L,

  • (1) – an elevated TSH concentration with a normal FT4 concentration,
  • (2) – an elevated TSH concentration with normal concentrations of FT3 and FT4,
  • (3) – an elevated TSH concentration with normal concentrations of TT3 and TT4, 35 obs. not shown
  • (4) – an elevated TSH concentration with normal concentrations of FT4 and TT4
  • (5) – an elevated TSH concentration with normal concentrations of TT3, TT4, FT3, and FT4.
5Crit 56297165885682507016170865
6Crit 2629716588566877682507016170865
7crit 162971658856687768250697047016170865
7Crit 462971658856687768250697047016170865
Table 1 – ID numbers and criteria

The results shown in table 1 indicate that criteria 5 is the most precise.


IDAGEFT3xFT4TSHTT3TT4x
166877804.15817.74.71147.1041
269704306.9313.74.663112.3551
Table 2 The T3 values in bold are outside the Chinese reference interval

That I have applied these criteria to the NHANES figures (Ref. 2) is of course not reasonable nor science but an attempt to illustrate what they are capable of. To me, it revealed that the reference interval is floating.

Next, I applied the reference intervals from NHANES to the 5 criteria.

The NHANES reference values as a basis – the difference between criteria 1 (28 obs) and criteria 5 (21 obs). The T3 values are important for judging subclinical values better than just for TSH/FT4 values. In table 3 are the 7 persons that differ between criteria 1 and 5.

IDFT3FT4TSHTT3TT4
701194.010.737.551176.83
700452.670.6112.49925.44
674601.730.7911.28698.24
660562.210.85.93667.9
651452.290.8842.09647.93
631503.310.856.151235.63
629192.980.777.241005.98

Ref. val
2,5 – 3,90.6 – 1.6
80 – 2006.09 – 12.23
Table 3 – ID numbers and characteristics of combinations of thyroid hormones (Those with TSH > 10 should be excluded – not SCH)

Four of the 7 not excluded would have been found by the FT3 value and only – three of the values would have been found by TT4. This follows the results seen in the Zheng et al article.

5 criteria plus TSH above the upper reference level gives the best result.

There has been a focus on the TSH levels as the indicator of subclinical hypothyroidism if FT4 at the same time in the reference interval. The summary of the FT4 values in criteria 5 is not sufficient in identifying the last seven persons. Only applying the FT3 reference interval identifies the last persons with subclinical hypothyroidism.

  1. https://doi.org/10.3389/fendo.2022.820414
  2. https://wwwn.cdc.gov/Nchs/Nhanes/Search/DataPage.aspx?Component=Laboratory&Cycle=2011-2012

Which doctor do you prefer?

There is an ongoing debate on the quality of life of hypothyroid patients. Some 10% of patients report hypothyroid symptoms although they are pharmacologically compensated.

Doctors may have different treatment goals:

When the patients TSH is in the green range they are satisfied with the treatment result.

Others choose a more refined treatment goal:

Seeking the exact thyroid profile that suits the patient.

Patients are individual and so are their TSH values.

Using another image that has been changed logarithmically shows how few persons you may find in each of the bars:

These images show how individual the TSH values are and that you should not accept an unsuccessful treatment. We have not yet looked at the two important hormones T4 and T3 but I will promise you that they will add to the individuality.

For now two tables with the raw data for the different figures that we will touch upon:

NHANES data

Thyroid Profile in this page – https://wwwn.cdc.gov/Nchs/Nhanes/Search/DataPage.aspx?Component=Laboratory&Cycle=2011-2012

Based on the data the following histogram was created showing counts for different levels of TSH.

Based on the data the following 3D scatterplot was created showing the high individuality of thyroid hormone combinations.

Running a script to identify distinct values showed that TSH, FT3 and FT4 were unique for 1964 observations.
For TSH, TT3, TT4 1949 observations were unique.

This shows based on these NHANES data, that the combination of the thyroid hormones is individual. Should treatment be based on averaged values or should the aim be finding the patients previous set-point?

Inspired by

doi:10.1530/EC-150056

Variation in the biochemical response to L-thyroxine therapy and relationship with peripheral thyroid hormone conversion efficiency

I have illustrated the SPINA-GD values from the NHANES data:

Then I looked at the FT3 and FT4 values also:

The two curves show changes perhaps reflecting the different categories from the SPINA-GD histogram.