The significance of the TSH-FT3 circadian rhythm

In my earlier post about circadian rhythms in thyroid health, I asked two quiz questions about TSH, FT3 and FT4 variability.

First I asked “Which value varies the most during the day, a) TSH, b) Free T3 or c) Free T4?”

  • There, I showed you, through various visuals by Russell et al, 2008, that TSH varies the most widely over a 24 hour period.

Then I asked: “Which thyroid hormone’s fluctuation is more variable from hour to hour, a) Free T4 or b) Free T3?”

  • I answered that both FT4 and FT3 are almost equally variable as percentage of the mean, and FT4 varies far more than FT3 in its absolute number of pmol/L. Again, I provided color-coded graphs by Russell et al, 2008 to show what the variation looks like.

I used other research to show you how this variability contributes to a regular, constant pattern of lab history over weeks or months, and how the stability of FT3 and FT4 can prevent misdiagnosis due to the TSH laboratory test timing.

Now I have a final quiz question for you.

Question #3. Which hormone’s circadian rhythm is more closely tied to the huge sine wave of TSH every 24 hours: is it a) Free T4 or b) Free T3?

If you chose b) Free T3, you have correctly answered the most important question for thyroid hormone health.

Only the Free T3 hormone’s 24-hour sine wave, not Free T4, has a strong echo of the nightly TSH rise in a person with a healthy thyroid and pituitary gland. The FT4 has a delayed wave and a flatter rise and fall.

The graphs by Russell and team reveal this pattern clearly, and I’ll show a couple of them again to begin.

I’ll comment further by using other research showing its biological significance in

  1. Human longevity
  2. Subclinical hypothyroidism

Why does this matter to you?

  • If you have some partial thyroid function remaining, you may be able to see these natural patterns break through your body’s response to hormone dosing.
  • If you know you have no thyroid function, you can interpret lab results more intelligently by ruling out the characteristic feedforward thyroid response, and
  • You can think about how dosing can compensate for an aging or failing thyroid or pituitary TSH.

Circadian rhythm patterns

As you can see in the way I’ve overlaid and colored Russell’s graphs, TSH (gray) clearly mapped onto the curve of FT3 (pink) more than FT4 (blue).


“There was a strong correlation between the time-adjusted FT3 and TSH levels.”

In contrast, the researchers found “no strong temporal relationship between FT4 and TSH with the peaks being spread quite uniformly.” In other words,

“The scatter plot revealed a weak correlation between FT4 and TSH.”

Feedforward is rhythmic

Russell and team concluded something significant from the TSH-FT3 circadian rhythm. They reasoned that

“The strong temporal relationship between TSH and FT3 levels and the positive correlation between time-adjusted FT3 and TSH levels suggest that the variation in FT3 levels is determined by TSH.”

This theory makes sense given what other researchers have found.

At the molecular level, a rise in TSH stimulates an increase in thyroidal T3 synthesis rate, which is boosted to a higher degree than the synthesis of T4 (Citterio et al, 2017).

Added to the TSH-enhanced T3 secretion rate is a degree of boosted T4-T3 conversion within thyroid gland tissue (Hoermann et al, 2013).

Of course, the rate of T4 secretion is also raised in response to TSH stimulation, but given the much higher level of Total T4 than Total T3 in blood, it is easier for an increase in T3 secretion to make a difference in blood levels in the short term. Therefore, the FT4 has relatively less immediate response to the large TSH rhythm.

The nightly dose of T3 from the pharmacy in the neck

Let’s look again at the individual example from Russell, the one with the strong TSH-FT3 rhythm.

Consider a pharmaceutical analogy. The nightly rise of TSH gives every person with a healthy thyroid gland a secret dose of T3 hormone overnight. TSH orders a T3 refill from the thyroid pharmacy in their neck. The TSH level is a prescription. It requests T3 at a time when the body needs it.

It is a relatively secret overnight T3 dose because the extra Free T3 in blood is mostly gone by 8 or 9am. The nightly dose provides a T3 “top-up” injected onto a stable foundation of T4 supply that is continually converting to T3 at a variable rate as TSH fluctuates.

The nightly T3 injection, over the long term, produces a steady product: an enhanced Free T3:T4 ratio the more the Free T3 is boosted every night in relation to Free T4 by means of a TSH rhythm.

Notice how both the FT3 and FT4 fall together in the first half of the day, but for most of the 24 hour period, FT3 is above the FT4 line in their respective reference ranges.

Therefore, the FT3:FT4 ratio, in light of the concurrent TSH level, can act as a gauge of the power of a person’s nightly feedforward TSH stimulation.

It reveals the T3-centric “feedforward” mechanism in action on a daily basis, whereas the counterbalance of T4-centric “negative feedback” on TSH operates on a larger time scale.

Nightly TSH-boosted FT3 levels associate with a longer life

Russell’s research was not capable of commenting on the health outcomes of daily circadian rhythms, but other research provides insight.

Researchers in 2015 (Jansen, et al) wnated to understand if anything was special about the thyroid health of the offspring of people who lived past age 90 (“nonagenarians”).

To target the nonagenarian gene profile more specifically, they selected only the offspring of a parent who lived past age 90 who had also had a brother or sister who also lived past 90. The spouses of these nonagenarians’ offspring served as the “partner” control group, as a contrast. They tested their TSH, FT3 and FT4 every 10 minutes for 24 hours.

They found that the nonagenarians’ offspring maintain a higher TSH and slightly higher nocturnal FT3 surge than the partner control group.

“during the nocturnal surge, the temporal relationship between TSH and fT3 was significantly stronger in offspring than in partners.”

The offspring of a longer-lived parent achieved this FT3 bonus not by means of a more exaggerated TSH rhythm, but by a TSH secretion that is overall higher in reference range, with an equally robust circadian rhythm, on average.

Their FT4, however, was equal to their spouses who are not children of 90-year-olds.

The end result of this raised TSH setpoint, nightly feedforward FT3 response, and equal FT4 is an elevated Free T3:T4 ratio, visible on their smaller line graphs.

See the significance of FT3:FT4 ratio in longevity

That small graph in the lower right corner, Figure D, is like the humble little clue to the big picture.

Jansen’s research team went a long way, but just could not take the next step to understanding because an old, distorted saying held them back.

Within the same year that they published this TSH-emphasizing graph, 2015, Jansen’s team produced a second article analyzing the same data set’s TSH-FT3 interaction over time. This article was titled

“Familial Longevity Is Associated With Higher TSH Secretion and Strong TSH-fT3 Relationship”

Jansen’s research team had a difficult time understanding what was going on between TSH and Free T3 and Free T4 because they took as a basic premise the hackneyed misinterpretation of Pilo’s 1990 study: “One hundred percent of circulating T4 is secreted by the thyroid gland; however, only 20% of T3 is derived from this source (1).”

Jansen’s citation 1 is Russell et al, 2008, who cited Pilo without attribution when they echoed this statement in their introduction. Pilo’s 1990 study had indeed revealed a 20% average, but in reality the full range of daily thyroidal T3 secretion ranged from 6.5% to 42% of daily supply, and there was no central tendency in the data set (T3 secretion was not clustered around 20%).

Jansen’s analysis of the circadian TSH-FT3 response nearly falters in their Discussion section because they resist the idea that the thyroid could possibly be secreting more T3 than 20% of their daily supply. “FT3 is predominantly produced by peripheral conversion, whereas a small amount is secreted by the thyroid gland,” they reiterate, as if the thyroid is not permitted to provide a larger ratio or amount.

Nevertheless, their previous studies had found no differences in peripheral conversion rate on average, so they had to concede (intelligently) that

“it seems likely that changes in fT3 relate to TSH stimulation of thyroid hormone release.”

Yet Jansen’s team believed the ability of the offspring for the thyroid to respond quickly, without delay, was the only clear difference, and that a small amount of extra T3 per unit of T4 at almost all times was not worth trumpeting.

Indeed, the time lag in FT3 response to TSH had been noted by Russell in 2008. After they removed “profiles with a time shift exceeding 12 h” they found “a close relationship between TSH and FT3 with 20 of 24 subjects showing peak correlation at between -0.5 and -2.5 h, suggesting that FT3 lags behind TSH by these amounts.” Years after Russell’s article, Jansen revealed that a sluggish FT3 response to TSH was not associated with longevity.

But what is the whole purpose of the timely FT3 response, if not to ensure that the FT3:FT4 ratio remains afloat at all times?

They simply misjudged the mildly elevated FT3:FT4 ratio in blood as physiologically insignificant just because it was statistically insignificant.

But physiology trumps statistics, especially when we are talking about the most essential and powerful thyroid hormone, T3.

The TSH-FT3 feedforward push in subclinical hypothyroidism

Just as an aging thyroid gland can be challenging, in untreated subclinical hypothyroidism, keeping the thyroid going as long as possible is challenging.

The conventional diagnosis of this gray zone of “subclinical” hypothyroidism is to have

  1. a TSH that is elevated but not above 10,
  2. in conjunction with FT4 hormone within reference range.

Its diagnosis as “subclinical” already requires two of the three hormones for correct diagnosis.

Measuring FT3 can aid diagnosis further, regarding whether this is “compensated” hypothyroidism (compensated by a higher FT3 and/or FT4) or “uncompensated.”

The TSH-FT3 circadian rhythm portrayed by Russell in 2008 and Jansen in 2015 is a key player in the body’s ability to compensate for as long as possible.

As recently as 2020, Hoermann and team have revealed how the TSH can be a savior in early thyroid failure. They pointed at what is going on with the TSH:FT3 ratio and the FT3:FT4 ratio to keep patients asymptomatic as long as possible.

As FT4 drops, TSH rises for a key purpose — to keep FT3 afloat around the midpoint of reference range, until the thyroid gland has lost too much of its function.

Some patients in this zone are completely asymptomatic when their TSH is elevated because TSH performs a positive feed-forward role if their thyroid gland is still healthy enough to respond mid-range to high-normal thyroid hormones.

An excellent study by Karmisholt et al in 2008 provides a complementary image that includes the FT4 hormone.

They examined the apparent stability of FT3 and FT4 and TSH in subclinical hypothyroidism over 13 months. They admitted 21 people to the study, but

  • one became euthyroid,
  • one became overtly hypothyroid,
  • and six progressed and were no longer stable.

This left the data set with 15 representations of subclinical hypothyroidism in which FT3 and FT4 were still very stable.

As you can see, the FT3 (in the red shade) is being kept afloat, but each individual has their own unique “fingerprint” of TSH, FT4 and FT3.

One can group together at least two types of fingerprints:

  • Patients 1, 2, 5, 7 and 15 were no longer merely “subclinically hypothyroid” — they were at times “hypothyroid” by the standard definition by FT4 (green dots) and TSH, but measuring their FT3 showed that it compensated. Patients 2, 5, 7, and 15 have the characteristically high FT3:FT4 ratio seen as FT4 drops lower and lower before the thyroid fails.
  • In contrast, the patients with the higher FT4 and FT3, especially 8, 9, and 10, have strong FT4 and FT3, and their FT4 is quite robust, keeping up with the FT3 or exceeding it.
    • It’s likely that their thyroids are responding well to a strong nightly TSH rise, not just a higher overall TSH level, because their TSH was not higher than others.
    • However, it looks very strange and contradictory for the TSH to be elevated when the thyroid hormones are so high, unless you understand that in these patients, the feedforward principle is overwhelming the feedback.

Karmisholt et al came to these observations and conclusions about their own research:

  • “Variation in TSH increased with increasing thyroid failure, while variation in thyroid hormones was unaltered” in each patient over the 13 months.
    • Patients 1, 6, and 15 show how the TSH dots disperse as the thyroid hormones fall lower and the elevated Free T3:T4 ratio is lost, and/or the hormones drop lower than the individual’s setpoint.
  • “To guide evaluation [of progression to thyroid failure], we found that a 40% difference was required between two test results to indicate a true difference for TSH in the range between 4.4 and 12 mU=L.
  • Due to more narrow variation in fT4 and fT3, the difference required for similar changes in fT4 and fT3 was approximately 15%.”

In other words, the TSH was not so “exquisitely sensitive” in subclinical diagnosis and monitoring after all.

  • Smaller changes (15%) in FT3 and FT4 within reference range revealed more of a pattern of thyroid hormone stability OR an impending crisis,
  • But a waffling elevated TSH had to rise by quite a bit (40%) before it truly signified a breakdown.

To prevent misdiagnosis, test more than TSH.

The results from these studies point to the importance of testing more than TSH, and the reliability of testing FT3 and FT4 to discover how well the thyroid is responding to circadian “feedforward” rhythms.

In 2016 Hoermann and team found that “univariate” diagnosics (relying only on one hormone) were far more inaccurate than “multivariate” thyroid diagnosis (a calculation relying on TSH, FT4 and FT3):

“Of the thyroid dysfunctions that were established by the univariate combination such as TSH alone, 26% (42%) were reclassified to “euthyroid” by using the composite bivariate (trivariate) reference method.”

They concluded that

“The classification of the thyroid disease should not rest on a single laboratory parameter, but rather consider the interactive involvement of all three parameters [TSH, FT3, and FT4].”

(Hoermann et al, 2016)

Conclusion: The trio of hormones

An extremely high TSH is always a clear signal of disease. Likewise, an extremely low TSH outside of thyroid therapy warrants further investigation.

However, many cases of thyroid disease diagnosis, and puzzles of human longevity, require more sensitive discernment than TSH alone can provide.

The rate of misdiagnosis is high when these hormones, especially TSH, are interpreted as isolated numbers within population-wide reference ranges. But we can significantly reduce diagnostic error by interpreting TSH, FT3 and FT4 together.

Why can’t it just be simple? Why can’t TSH tell us the whole story?

  • Because the TSH population reference range is both too narrow and too wide to describe the highs and lows of robust circadian rhythm and early gland failure (thyroid and/or pituitary failure), and
  • Because it matters to human health how much TSH can rise, and how promptly the body responds to TSH’s nightly call with a timely FT3 dose.

The healthy human circadian rhythm of TSH is what makes FT3 such an important and insightful test of thyroid function, even while the thyroid is slowly failing or slowly aging.

The feedforward principle we see on a daily basis is a pattern that maintains homeostasis. Rhythmic patterns of overlapping sine waves, like a pumping heart, contribute to a narrow weekly and monthly stability of FT3, FT4 and TSH in health. This stability declines and fails in hypopituitary and/or hypothyroid diseases.

A boosted (or flagging) daily rhythm yields an FT3:FT4 ratio that even if small, can become diagnostic of a robust (or failing) response to TSH.

These three lab tests are an inseparable “trio” that can sing together in thyroid hormone harmony.

Circadian rhythms of TSH and FT3, in the context of FT4, provide increased discernment to diagnosis of subclinical hypothyroidism and can even provide clues to human longevity.

[Read part 1: Circadian Rhythms of TSH, FT3 and FT4 in thyroid health]


Citterio, C. E., Veluswamy, B., Morgan, S. J., Galton, V. A., Banga, J. P., Atkins, S., Morishita, Y., Neumann, S., Latif, R., Gershengorn, M. C., Smith, T. J., & Arvan, P. (2017). De novo triiodothyronine formation from thyrocytes activated by thyroid-stimulating hormone. The Journal of Biological Chemistry, 292(37), 15434–15444.

Jansen, S. W., Akintola, A. A., Roelfsema, F., van der Spoel, E., Cobbaert, C. M., Ballieux, B. E., Egri, P., Kvarta-Papp, Z., Gereben, B., Fekete, C., Slagboom, P. E., van der Grond, J., Demeneix, B. A., Pijl, H., Westendorp, R. G. J., & van Heemst, D. (2015). Human longevity is characterised by high thyroid stimulating hormone secretion without altered energy metabolism. Scientific Reports, 5, 11525.

Jansen, S. W., Roelfsema, F., van der Spoel, E., Akintola, A. A., Postmus, I., Ballieux, B. E., Slagboom, P. E., Cobbaert, C. M., van der Grond, J., Westendorp, R. G., Pijl, H., & van Heemst, D. (2015). Familial Longevity Is Associated With Higher TSH Secretion and Strong TSH-fT3 Relationship. The Journal of Clinical Endocrinology and Metabolism, 100(10), 3806–3813.

Karmisholt, J., Andersen, S., & Laurberg, P. (2008). Variation in thyroid function tests in patients with stable untreated subclinical hypothyroidism. Thyroid: Official Journal of the American Thyroid Association, 18(3), 303–308.

Hoermann, R., Midgley, J. E. M., Larisch, R., & Dietrich, J. W. (2013). Is pituitary TSH an adequate measure of thyroid hormone-controlled homoeostasis during thyroxine treatment? European Journal of Endocrinology, 168(2), 271–280.

Hoermann, R., Pekker, M. J., Midgley, J. E. M., Larisch, R., & Dietrich, J. W. (2020). Triiodothyronine secretion in early thyroid failure: The adaptive response of central feedforward control. European Journal of Clinical Investigation, 50(2), e13192.

Russell, W., Harrison, R. F., Smith, N., Darzy, K., Shalet, S., Weetman, A. P., & Ross, R. J. (2008). Free Triiodothyronine Has a Distinct Circadian Rhythm That Is Delayed but Parallels Thyrotropin Levels. The Journal of Clinical Endocrinology & Metabolism, 93(6), 2300–2306.

Categories: Healthy thyroid axis, Thyroid Testing

3 replies

  1. Thank you for such an informative article. If I interpreted this correctly, for the hypothyroid patients on combined T3+T4 therapy, it would seem that taking the T4 dose early in the day and taking most of the daily T3 dose in the evening would help the body match the normal circadian rhythm of these hormones.

    • Good question, Julien. Thyroid therapy is never going to totally mimic the natural circadian rhythm, but we can certainly try.

      This is going to be a long answer because these posts focused on the “healthy thyroid” and people who are not yet in need of interfering with natural rhythms by using thyroid medication. I think I’ll make it a new post.


  1. Q&A. Dosing T3 in light of circadian rhythm – Thyroid Patients Canada

Leave a public reply here, on our website.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: