Published Research on T3/T4
as a Treatment for Mood Problems
(Updated Jan 2004; slight revisions July 2004; reread for accuracy 4/2008)

Here are a the few studies that directly address use of T3 with T4, with a "translation" or comment after each.   For comparison, you may be aware or become aware of the great energy already on the Internet regarding the use of T3 as a miracle cure, for depression and fibromyalgia in particular.  There is almost certainly more to this than "placebo", but the only way we'll find out for sure is to compare against a placebo treatment.  This has not happened yet.  Therefore I would suggest interpreting the work of Shomon, Lowe, Arem and others with caution for now -- or anything I say, for that matter.  We know too little now to be declaring a "thyroid solution".  At least I know too little for that, and so far can't find much basis for certainty about thyroid as a central element in treatment of mood disorders.  In part for your amusement, I've included below a sample of the energy around this issue. 

Here are the studies that have been published in medical journals listed in the National Library of Medicine (that includes all the major journals and many minor ones), from a literature search in late July 2001. 

[Update Jan. 2004:  Studies "Five and Six" below are the largest, and represent an attempt to repeat the big study that really got a lot of interest going about T3/T4 combinations (Study 4 below, Bunevicius, NEJM 1999).  Some authors would recommend starting with the last two, numbers 5 and 6,  rather than going back through the history, because they contradict the main findings of the earlier studies and are based on a much larger sample size and/or better study design. ]

Study 1. 9 patients, no placebo

Cooke RG, Joffe RT, Levitt AJ.   T3 augmentation of antidepressant treatment in T4-replaced thyroid patients.  J Clin Psychiatry 1992 Jan;53(1):16-8

BACKGROUND: Clinicians may not consider using the thyroid hormone liothyronine sodium (levorotary isomer of triiodothyronine [T3]) for augmentation of antidepressant drugs in depressed patients who are also receiving the precursor hormone levothyroxine (levorotary isomer of thyroxine [T4]) for thyroid disease. We now report on the successful use of T3 augmentation therapy in seven of nine depressed patients who were also receiving T4 for thyroid disease. METHOD: Following an earlier single case report, we prescribed T3 augmentation therapy for eight depressed patients who had not responded to an adequate antidepressant drug trial and who were receiving T4 therapy for thyroid disease. T3 was prescribed in open-label fashion, and response was judged by the clinician, whose assessment was supplemented by the use of standardized rating scales. RESULTS: Seven of the nine patients were judged to respond to T3 augmentation. CONCLUSION: These results are consistent with a report of differential effects for T3 versus T4 augmentation in depressed patients free of thyroid disease. The results have implications for the treatment of depression in the presence of thyroid disease and for the mechanism of thyroid hormone potentiation of antidepressants. (italics mine)

Translation: The authors tried adding T3 in nine people who were already on T4.  These patients were all having symptoms of depression, and all taking antidepressant treatment.  They were on T4 because of known thyroid problems, which were identified separately from the depression problems.  When T3 was added to the T4 they were already receiving, seven of the nine patients (78%) clearly became less depressed.

Study 2. 1 patient, no placebo -- a "case report"

Rack SK, Makela EH.  Hypothyroidism and depression: a therapeutic challenge.
Ann Pharmacother 2000 Oct;34(10):1142-5.

OBJECTIVE: To describe a patient with longstanding depression and hypothyroidism who had marked mood improvement only after triiodothyronine (T3) was added to her thyroxine (T4) replacement therapy. CASE SUMMARY: A 50-year-old white woman had a long history of depression and documented hypothyroidism since 1991. Despite treatment with T4 with dosages up to 0.3 mg/d, she continued to be depressed, have symptoms of hypothyroidism, and have a persistently elevated thyroid-stimulating hormone concentration. Addition of a low dose of T3 to her regimen resulted in significant mood improvement. DISCUSSION: The relationship between hypothyroidism and depression is well known. It is possible that this patient's long history of depression may have been a consequence of inadequately treated hypothyroidism, due either to poor patient compliance or resistance to T4. Nevertheless, her depression responded to addition of a low dose of T3 to her regimen. This case emphasizes the importance of screening depressed patients for hypothyroidism. Her clinical course also suggests that depression related to hypothyroidism may be more responsive to a regimen that includes T3 rather than to replacement with T4 alone. This is consistent with the observation that T3 is superior to T4 as adjuvant therapy in the treatment of unipolar depression. CONCLUSIONS: Depressed patients should be screened for hypothyroidism. In hypothyroid patients, depression may be more responsive to a replacement regimen that includes T3 rather than T4 alone. Therefore, inclusion of T3 in the treatment regimen may be warranted after adequate trial with T4 alone.  (italics mine)

Comment: only 1 patient, but we can add her to the 9 above.

Study 3. 26 patients, no placebo

Bunevicius R, Prange AJ.   Mental improvement after replacement therapy with thyroxine plus triiodothyronine: relationship to cause of hypothyroidism. Int J Neuropsychopharmacol 2000 Jun;3(2):167-174

We treated 26 hypothyroid women - 11 with autoimmune thyroiditis and 15 who had been treated for thyroid cancer - with their usual dose of thyroxine (T4) or with a regimen in which 50 mcg of T4 had been replaced by 12.5 mcg of triiodothyronine (T3). Patients were first randomly assigned to one regimen for 5 wk and then to a second regimen for an additional 5 wk. The substitution of T3 for a portion of T4 caused expected changes in concentrations of thyroid hormones and thyroid-stimulating hormone (TSH). After combined hormone treatment there were clear improvements in both cognition and mood, the latter changes being greater. The patients who had been treated for thyroid cancer showed more mental improvement than the women with autoimmune thyroiditis, perhaps because they were more dependent on exogenous hormone. Some mood improvements correlated positively with changes in TSH while others correlated negatively with changes in free T4.  (italics mine -- JP)

Comment:  This research design offers the opportunity to directly compare T4 alone with T4/T3 combined, but there is still no placebo group, so it is still basically a collection of cases similar to the nine cases above and to my list of cases.  You can see that the authors again found that T3 and T4 seemed more effective for mood than T4 alone.

Study 4. Most of these 33 patients are the same ones reported on in the previous paper above, I believe; so this reports an additional 7 cases: 

Bunevicius R, Kazanavicius G, Zalinkevicius R, Prange AJ Jr.  Effects of thyroxine as compared with thyroxine plus triiodothyronine in patients with hypothyroidism.  N Engl J Med 1999 Feb 11;340(6):424-9

BACKGROUND: Patients with hypothyroidism are usually treated with thyroxine (levothyroxine) only, although both thyroxine and triiodothyronine are secreted by the normal thyroid gland. Whether thyroid secretion of triiodothyronine is physiologically important is unknown. METHODS: We compared the effects of thyroxine alone with those of thyroxine plus triiodothyronine (liothyronine) in 33 patients with hypothyroidism. Each patient was studied for two five-week periods. During one period, the patient received his or her usual dose of thyroxine. During the other, the patient received a regimen in which 50 microg of the usual dose of thyroxine was replaced by 12.5 microg of triiodothyronine. The order in which each patient received the two treatments was randomized. Biochemical, physiologic, and psychological tests were performed at the end of each treatment period. RESULTS: The patients had lower serum free and total thyroxine concentrations and higher serum total triiodothyronine concentrations after treatment with thyroxine plus triiodothyronine than after thyroxine alone, whereas the serum thyrotropin concentrations were similar after both treatments. Among 17 scores on tests of cognitive performance and assessments of mood, 6 were better or closer to normal after treatment with thyroxine plus triiodothyronine. Similarly, among 15 visual-analogue scales used to indicate mood and physical status, the results for 10 were significantly better after treatment with thyroxine plus triiodothyronine. The pulse rate and serum sex hormone-binding globulin concentrations were slightly higher after treatment with thyroxine plus triiodothyronine, but blood pressure, serum lipid concentrations, and the results of neurophysiologic tests were similar after the two treatments. CONCLUSIONS: In patients with hypothyroidism, partial substitution of triiodothyronine for thyroxine may improve mood and neuropsychological function; this finding suggests a specific effect of the triiodothyronine normally secreted by the thyroid gland.

Comment: This was the study that got things rolling on T3.  Note the authors measured mood and thinking ability ("cognitive performance") and found improvement in both.  What we need now is a placebo-controlled study, where the T3 is added in some, but not in others, and neither the patients nor the researchers know got it and who didn't.  [That study arrived in 2003, shown next.]

Study 5.  40 subjects, 20 of whom were left on T4 alone while 20 were given T3 substituting for some of their T4

Sawka, A M.; Gerstein, H C.; Marriott, M J.; MacQueen, G M.; Joffe, R T.  Does a Combination Regimen of Thyroxine (T4) and 3,5,3'-Triiodothyronine Improve Depressive Symptoms Better Than T4 Alone in Patients with Hypothyroidism? Results of a Double-Blind, Randomized, Controlled Trial.  Journal of Clinical Endocrinology & Metabolism, Volume 88(10), October 2003:  4551-4555.

Study 6.  46 subjects, 23 in each group:  T4 plus a placebo, or T4 plus T3.  

Clyde et al.   Combined levothyroxine plus liothyronine compared with levothyroxine alone in primary hypothyroidism: a randomized controlled trial.  JAMA. 2003 Dec 10;290(22):2952-8.

The authors of studies 5 and 6 conclude that there is no benefit from adding T3 to T4.  Their studies were very well designed and conducted.  Perhaps they are right.  Perhaps there are only a few people who benefit from this approach -- like some of my patients seem to -- and they were lost in the group data in this study.  Or perhaps there really is a benefit, but somehow it just didn't show up here.  Is there anything in the study data that might explain why a benefit, if it really exists, didn't show up? 

Here' s one:  if you look at the TSH and T4 and T3 levels during study 5, you see that the TSH was held pretty close to 1-1.5, a pretty ideal level, through the experiment.  But in the T3-added group, where they lowered T4 to "make room" for the T3 (so as not to lower the TSH too much), there was quite a drop in T4 levels compared to the stay-on-T4 group; and there was much less of an increase in T3: 

Maybe this difference in T4 replacement means something, maybe not.  Maybe there just isn't any benefit from using T3 with T4 in some people and I've allowed myself to be tricked into thinking so (based on my experience with a few patients, and the data suggesting that stressed people can't make TSH as well or convert T4 into T3 as well.Tsigos, Charmandari ).  I will still use T3/T4 combinations in some patients, but especially after studies 5 and 6, I no longer use it first, before T4 alone, when someone becomes hypothyroid on lithium, for example.  

Samples of the Endocrinology discussion of T4 versus T3 and T4 together

There is quite a controversy in a small niche in endocrinology between the old guard who believe that T4 is the only replacement to consider using, and a small little group who think that making TSH normal using T4 may still leave some people effectively hypothyroid.  The latter group think that people may have trouble making TSH in the first place and so can look "over-replaced", i.e. low TSH, not because they have too much thyroid but because they just can't make TSH in the first place.  So they look hyperthyroid, but may actually be hypothyroid.   This latter group often agree with a small group of patients (maybe not so small anymore, as this discussion gets aired on the internet) that doctors should pay more attention to how patients feel, and what they say about how they feel, than what the TSH is. 

For an example of this discussion, forwarded to me by one of my patients who follows these things very closely (thank you), here are two such views.  First,  a retired British doctor gives the "treat the patient, not the lab value" point of view in a letter to the British Medical Journal.  Then I'll give you the link to the editorial to which he refers, a classic version of the old-guard view, complete with tone of voice, as the first doctor points out. 

Dear Sir

The editorial in the BMJ 2003, 8th February, by Toft and Beckett is disappointing in a number of regards. The opening paragraph encapsulates the generalised and arrogant approach taken by Dr Toft together with so many of his colleagues. His suggestion that patients with symptoms of hypothyroidism but normal blood tests, are a “vociferous minority” who have psychosocial problems, is entirely misleading and quite insulting to a great many patients. If he could be persuaded to take the trouble to undertake a full clinical appraisal, including the invaluable basal temperature test, suggested by Barnes as long ago as 1945 in the Lancet, he would doubtless be able to make a diagnosis of hypothyroidism on clinical grounds, and learn for himself the pitfalls of an uncritical reliance on blood tests.

His second paragraph, concerning “misguided medical practitioners” who in their alleged unwisdom actually try to solve their patients’ problems by using T3 as well as T4, or natural desiccated thyroid, is again gratuitously insulting. These practitioners do not attempt to hit some desired T4 or TSH level, but actually titrate the dosage to their patients’ needs. A process called listening to the patient. Reliance on the “sensitivity” of the TSH, expressly disregarding the clinical response, is misplaced; the test approach must have a sensitivity of at least 0.02 mu/l, which is difficult to reliably achieve and thus affords a frequently seen cause for error.

He raises again the spectre of iatrogenic hyperthyroidism as a cause of osteoporosis and atrial fibrillation. Although a theoretical possibility, there is no evidence for this and I have never seen it in many thousands of patients. It is particularly unlikely when patients have been taught the value of self-monitoring.

Dr Toft refers to Pollock et al, who purported to demonstrate, in a much criticised paper, that thyroxine helped a little in patients with incontrovertible hypothyroidism but didn’t at all if they hadn’t. As with the authors, Dr Toft confuses symptoms with a full clinical appraisal providing a consequent diagnosis, which it is perfectly proper for an experienced physician to make despite a normal TSH and T4. Neither this paper nor Dr Toft’s comments have furthered understanding in any way.

Again, he refers to “the exquisite” sensitivity of the TSH which leads him and his colleagues astray. He is right to point out that pituitary or hypothalamic weakness, very much more common than he believes, will provide a false low TSH; he would do well to remember that a hypometabolic pituitary/hypothalamic axis is a commonly met cause of unreliable TSH estimations.

However his observation that a replacement providing a higher than normal T4 and suppressed TSH allows for greater patient well-being is to be welcomed. Further is to be welcomed his consideration concerning the addition of T3 to T4, if somewhat rather late in the day. Perhaps he might in due course be persuaded that the use of T2 as well, as with desiccated thyroid, confers additional advantage.

Yours faithfully

Dr Barry Durrant-Peatfield M.B., B.S., L.R.C.P, M.R.C.S.

Editorial to which Dr. Durrant-Peatfield refers


Relevant additional studies or articles

This is just a file for my collection of interesting implications on this subject:

(this slide is the work of, and shown courtesy of, a gracious thyroid researcher who requests not to be named)