Another great study by McAninch and colleagues, from Antonio Bianco’s lab, has been published in a leading endocrinology journal, The Journal of Clinical Endocrinology & Metabolism.
The study was announced and summarized in this press release titled “Standard Hypothyroidism Treatment Falling Short.”
Let’s hope the endocrinology community will listen and think twice about whether its guidelines are working for all patients.
Below, I supplement their press release with my own analysis of the full article.
McAninch, E. A., Rajan, K. B., Miller, C. H., & Bianco, A. C. (2018). Systemic Thyroid Hormone Status During Levothyroxine Therapy In Hypothyroidism: A Systematic Review and Meta-Analysis. The Journal of Clinical Endocrinology & Metabolism. (advance online publication as of August 27, 2018)
In the article, they reviewed 99 prior articles to discover treated patients’ measures not just for cholesterol but for kidney, liver, ferritin, cognitive health, etc.
Shortcomings in thyroid research
Unfortunately, the research team often found the studies difficult to synthesize because some studies did not compare with healthy control subjects, different studies used different methods, and some important measures just had not been sufficiently studied in the literature.
For example, on cognitive health:
There were seven studies identified in the search that met inclusion criteria and assessed cognition in healthy controls and LT4-treated hypothyroid patients.
Although different cognitive tests were performed in these individual studies, there was only one study that identified any cognitive test to be worse among LT4-treated hypothyroid patients than healthy controls, which found the visual scanning test to take significantly longer in treated hypothyroid patients. (22)
Overall, our interpretation was that these objective cognitive measures were normalized in LT4-treated hypothyroidism at doses that normalize the serum TSH, but this result is limited by the heterogeneity of cognitive tests utilized in these individual studies.
Twelve studies that measured cognition in LT4-treated participants, but not healthy controls, were identified. It was difficult to cohesively compare these studies
and draw conclusions because of (i) heterogeneity of cognitive tests employed, and (ii) lack of stated reference ranges for these tests in most of the studies. Notably, in the studies that did specify reference ranges for the cognitive tests utilized, (30-32) the outcomes were normalized by LT4 treatment.
(from Results: Cognition section)
What is going on with research on therapy for hypothyroidism?
They looked at 99 articles and only seven studies even asked this question, and only one study identifies the continued symptoms that so many patients experience on therapy.
Millions of people are on this therapy and the thyroidologists are focusing on the rarer thyroid conditions and cancers … yet those patients may eventually end up on therapy for hypothyroidism too.
Is it perhaps because the assumption is that “we’ve fixed hypothyroidism”? So there is not enough doubt about the effectiveness of the preferred therapy and the standard guidelines? It is not worthy of careful systematic study to ensure that it is, indeed, “fixed” for everyone?
Back to the article’s findings.
Cognition and hypothyroid therapy
Here’s some of the challenges that come with this particular area of research.
When you are studying cognitive health in thyroid therapy, it can be beneficial to design or adopt your own “cognitive test” because other tests used by previous studies may not be specific to the types of cognitive changes likely to be experienced in hypothyroidism or hyperthyroidism.
If you want to be sure to include the right measures to test your hypotheses on thyroid hormone effects, as a researcher you have to first understand how thyroid hormone signalling affects the brain. You might have to look at research done many decades ago, since who is studying these basic questions anymore? You then combine this physiological knowledge of thyroid hormones with clinical experiential knowledge of treating a variety of hypothyroid patients with cognitive difficulties and seeing how changes in their thyroid dosage affects their symptoms. Then you can develop a good hypothesis about how thyroid hormone levels can directly or indirectly affect clinical symptoms, such as “visual scanning.”
The overall score on any cognitive test will be influenced by all the measures in your cognitive test, so if only a few of the indices are likely to be affected by thyroid status, the instrument will bury them under a larger pile of data — unless you carefully mine the pile of data to look for patterns within it. Interpretation of any data will rely on the researcher’s theoretical and clinical knowledge, assumptions and hypotheses.
Unfortunately, later, if other people do studies using completely different cognitive tests, there’s no way to compare their results across studies, because the measuring stick has changed.
Your study may end up being “one of a kind,” the only one that revealed a small aspect of so many treated thyroid patients’ mental and emotional difficulties.
With blood markers like ferritin or cholesterol, at least the measuring stick does not change as much as it does in psychology.
LDL and total cholesterol
The most significant finding was regarding cholesterol levels. Indeed, the data they discuss can reasonably lead one to suspect that not every patient’s cholesterol is sufficiently lowered on L-T4 without the addition of statin drugs.
Sadly, the article did not mention or cite references reminding people that it has long been established (since the 1920s) that cholesterol is directly, inversely tied to T3 levels even within the normal T3 reference range, and it happens through known biological mechanisms (Shin et al, 2003; Rizos et al, 2011, and Vivo Physiology n.d.).
Even though this effect has been known since the 1920s, it seems that many doctors, even younger endocrinologists, have never been taught this. Therefore, some uninformed readers may think that it’s not the job of effective thyroid therapy to lower cholesterol, that’s the job of a statin drug or your diet. But yes, LDL and total cholesterol actually are a good index of your thyroid hormone sufficiency and cholesterol levels, and higher cholesterol can correlate with symptoms (Guarnizo-Poma et al, 2018).
Thyroxine (T4) does not directly change cholesterol levels as much as triiodothyronine (T3) simply because T4’s effect on T3 levels depends on the a patient’s of T4-T3 conversion.
A sufficient T4 dose has to actually raise T3 levels in order to lower cholesterol, since T4 molecules do not affect LDL and Total cholesterol, but T3 molecules do (see Chehin, 1995).
As a result, a rise in T3 levels does not occur to the same degree in all people who take Levothyroxine, and the degree of “rise” depends on how low their T3 was to begin with.
Unfortunately, since the improvements to the TSH test (in the 1980s and 1990s) many studies on thyroid hormone therapy have tended to avoid measuring Free T3 and T4, assuming that the TSH tells all (Michalpolou, 1998).
Failing to measure Free T3 and failing to analyze Free T3 as a continuous variable (rather than just seeing if it is “in normal range” or not) are flaws in too many studies that try to investigate the connection between thyroid hormones and health. Why do researchers think it’s okay not to measure the most important and active thyroid hormone in such studies? Alas, the TSH dogma blinds people to T3’s independent and powerful effects.
Monotherapy vs. combination therapy
It’s good that McAninch et al. mentioned the fact that not everyone is able to achieve optimal T3 on LT4 monotherapy. They mention the well known fact that T4 therapy yields a higher T4:T3 ratio than normal, and they cite studies that show “combination” T4/T3 therapy is more likely to restore euthyroid T3 levels to peripheral tissues.
They had to admit that no proof exists that makes combination therapy “superior” to L-T4 for all patients. True.
My question is — why does T4/ T3 combination therapy have to be superior for everyone?
Are we in a competition to replace one therapy with another for ALL patients?
No we are not.
Patients require the therapy that works for their unique thyroid hormone health condition, and some people have genetic handicaps in thyroid hormone conversion, transport, and/or hormone receptor sensitivity.
If adding T3 hormone provides superior therapy for a significant subgroup, then that is clearly life-changing to that subgroup.
If some people convert T4 into sufficient T3, let them live happily with T4 monotherapy.
Why add T3 when the patient does not need extra T3? Give T3 to those who need it.
The article’s stance on normalizing TSH
The discussion section of the article was still very conservative about the importance of normalizing TSH, even when using T3 hormone in therapy. They recommend achieving a “normal” TSH on this therapy, as well as a stable T3 and T4:T3 ratios found in untreated healthy patients.
“Further clinical trials of combination therapy utilizing delivery mechanisms that provide stable serum T3 levels, and maintain physiologic serum T4:T3 ratios, while maintaining serum TSH within normal range, are justified.”
Unfortunately, not every patient can achieve optimal results by imitating the serum T3, T4 and TSH levels of untreated, healthy patients.
Thyroid therapy’s true goal is to restore health to the individual patient, not to recreate an idealistic ratio between hormones that imitates nature.
Some patients who are “poor converters” of T4 will require more T3 dosage in order to achieve optimal T3 levels, and this dosage will result not only in an unnaturally lower T4:T3 ratio but often a suppressed TSH. This is because T3’s oral dosing effect at larger doses will lower the TSH, even if Free T3 is not yet high enough within range to remove hypothyroid symptoms.
A lower TSH and a lower T4:T3 ratio than normal “normal” is a price well worth paying if that is what it takes for that individual to have a healthy life. (And no, there is no scientific proof yet that a low TSH by itself can harm you if your T3 is not too high at the same time (Rationale: Low TSH vs. true hyperthyroidism ; see also our blog post on Osteoporosis).
On the other hand, McAninch et al’s recommendations make sense for clinical research study designs. if TSH and T4:T3 ratios were “normalized” in all combination therapy studies, it would have the benefit of making them capable of being compared with L-T4 monotherapy studies and with each other, too.
Further Commentary: The aim of TSH normalization
It was not the aim or role of this article to address the thyroid hormone guideline of TSH normalization, but it is worthy of further commentary.
No, TSH is not an infallible index of euthyroidism in a person taking any sort of thyroid hormone therapy.
TSH is an index of the effect of thyroid hormones on the hypothalamus-pituitary complex, assuming full health within their tissues. We know that the hypothalamus and pituitary respond to T4 and T3 amounts and ratios differently than other glands and tissues in the human body (see Werneck de Castro et al, 2015).
This means that if the hypothalamus and pituitary are satisfied with T4 and T3 levels and secrete “normal” TRH and TSH, it does not mean that the rest of the brain and body is receiving enough T3. There’s a complex relationship even just between the hypothalamus secretion of TRH hormone and the pituitary gland’s response to TRH. The complex system makes it too easy for TSH to have interferences and abnormal responses.
In addition, the function of TSH shifts in thyroid disease, so when you measure TSH in a thyroid patient, you’re now measuring a relatively impotent hormone. This matters. It’s not a true feedback “loop” when the TSH-T4-T3 “loop” is broken by not having a fully functional gland for TSH to regulate. The signals that the pituitary sends out are not having their expected, normal effects on your body, so the pituitary is actually getting distorted feedback in return:
- A living thyroid gland adapts to supply relatively more T3 than T4 when TSH rises and relatively less T3 when TSH falls (Carpi, 1979).
- Your hypothalamus and pituitary gland can and still do respond to cold conditions, infections, and fasting (Fekete, 2007).
- However, no change in TRH and TSH secretion will make your hormone pill change increase or lower its dose, or change its T4:T3 ratio.
Consider patients with varying degrees of thyroid failure and their artificial, static doses of hormone replacement. In therapy, you can get pathological distortions in T4:T3 ratios (high-normal T4 with below-range T3 and a suppressed TSH). These distortions cannot be revealed by, much less modified by, TSH secretion.
Overall, TSH is a test that conveniently and cheaply identifies overt thyroid failure before therapy begins. That does not mean that “TSH normalization” can do all the work of optimizing Free T3 levels for the individual thyroid patient on therapy.
The inappropriate expansion of TSH from a population-wide diagnostic test to an individual therapeutic goal is well critiqued by Hoermann et al, 2016.
Analogy from a children’s story: Humpty Dumpty fell off the wall and broke into pieces. You can’t put him back together again perfectly by putting his pieces back on the wall.
Even if you can restore most of a patient’s health, they may always limp or may require a wheelchair from now on.
Admit it. All thyroid hormone therapy is artificial. It’s not always appropriate to apply the measuring-stick of perfect health that applies to the untreated population (TSH and normal hormone referernce ranges). It can only be an approximate guide or starting point, and it is not the ultimate judge of an effective outcome.