TCPR: When do you use thyroid augmentation in bipolar disorder?
Dr. Kelly: It may be useful for treatment-resistant bipolar depression, and I emphasize treatment-resistant because that is where the empiric support lies. Now, the evidence is not rock-solid, but when it comes to treatment-resistant bipolar depression we don’t have much else to draw on except maybe ECT. Thyroid augmentation has been tested in two double-blind, placebo-controlled trials. It worked in one, but the second trial found only a nonsignificant trend favoring thyroid (Walshaw PD et al, Bipolar Disord 2018;20(7):594–603; Stamm TJ et al, J Clin Psychiatry 2014;75(2):162–168).

TCPR: How do you explain the difference in the results?
Dr. Kelly: The positive study was done in rapid cycling bipolar disorder, and there is evidence of thyroid dysfunction in rapid cycling cases. But in my experience, it works in bipolar depression that is not rapid cycling as well. The negative study is considered a failed one because it had a high placebo response rate in men. It actually worked in women, and there may be something to that. There’s a trend for women to respond better than men in the thyroid augmentation research.

Table: Thyroid Augmentation in Bipolar Disorder

(Click to view full-size PDF.)

TCPR: Do we know how thyroid might work in bipolar depression?
Dr. Kelly: The leading theory is that people with bipolar disorder have difficulty getting thyroid hormone into their cells. This is speculation, but it is based on a well-replicated finding in bipolar disorder. People with bipolar disorder have low levels of adenosine triphosphate (ATP), and you need ATP to transport thyroid hormone across the cell membrane (Scaini G et al, Mitochondrion 2021;57:23–36). So even if they have normal levels of thyroid hormone circulating in their bloodstream, they may have hypothyroidism at the cellular level. By using higher doses of “supraphysiologic” thyroid, we push enough inside the cells to make up for the hypothetical deficit.

TCPR: Is this theory unique to bipolar depression, or does it apply to any depression?
Dr. Kelly: So far it seems limited to bipolar disorder—including bipolar I and II depression. Thyroid augmentation works in unipolar depression as well, but we don’t know as much about the mechanism there.

TCPR: How does thyroid augmentation affect mania?
Dr. Kelly: We don’t have much research on that. In my own experience with around 2,000 patients, I have not seen thyroid trigger mania. In the controlled trial of rapid cycling bipolar disorder, thyroid augmentation did not worsen mania, and it improved mixed states (Walshaw et al, 2018).

Risks

TCPR: What side effects do you warn patients about when you’re starting thyroid?
Dr. Kelly: Mainly I warn them about feeling too hot and an increase in anxiety and heart rate. I monitor actively for tachycardia and will decrease the dose if the pulse goes higher than 100 bpm at rest. I’ll also pay attention to patients’ caffeine use, because too much caffeine will raise the risk of tachycardia. I explain to them that thyroid can be very helpful for anxiety but that too much thyroid can cause anxiety—in which case we would need to lower the dose.

TCPR: That sounds counterintuitive. Thyroid augmentation helps anxiety?
Dr. Kelly: Yes. In one of the controlled trials, thyroid augmentation reduced anxiety in bipolar disorder, even in those who had a lot of anxiety at the start of the trial (Pilhatsch M et al, Int J Bipolar Disord 2019;7(1):21). That’s been true to my experience as well.

TCPR: Do any of these side effects potentially point to a serious hazard?
Dr. Kelly: Not if you lower the dose when they appear. Those side effects are signs that you’ve induced clinical hyperthyroidism, and that usually improves if you lower the dose.

TCPR: I’m surprised to hear it’s well tolerated.
Dr. Kelly: Thyroid is better tolerated in bipolar disorder than you’d expect. There’s an interesting study where researchers gave high-dose T4 (levothyroxine, Synthroid 500 mcg) to patients with refractory depression, both bipolar and unipolar, and to a healthy control group. The patients with mood disorders tolerated it well, with no dropouts, but 38% of those without mood disorders dropped out because they couldn’t tolerate the side effects, such as anxiety, tremor, restlessness, headache, tachycardia, and sweating (Bauer M et al, J Affect Disord 2002;68(2–3):285–294).

TCPR: What about medical risks of hyperthyroidism?
Dr. Kelly: The main risks are to the bones and heart. In theory, there is a risk of decreased bone mineral density and cardiac arrhythmias with hyperthyroidism. We actually don’t see those in the studies on mood ­disorders, even with long-term follow-up, but these studies probably are not powered to detect it (Ricken R et al, J Affect Disord 2012;136(1–2):e89–e94).

TCPR: But we do see osteoporosis and arrhythmias in patients with hyperthyroidism, right?
Dr. Kelly: Yes, that is true in clinical hyperthyroidism—when excess thyroid comes from the patient’s own glands, as in Graves’ disease. But there is some evidence that those problems are not due to the excess thyroid itself but to the autoimmune antibodies that are part of the disease. When the excess thyroid is exogenous—from a pill—we really don’t have good evidence that it causes bone or cardiac problems. For example, people who have had their thyroid removed after cancer are often put on high doses of thyroid hormone to suppress the TSH. We have a lot more studies in these patients than we do for high-dose thyroid in mood disorders, and the data are generally reassuring. High-dose thyroid does not seem to elevate the risk of osteoporosis, heart disease, or mortality (Kelly T et al, Prog Neuropsychopharmacol Biol Psychiatry 2016;71:1–6).

TCPR: What role do labs play in this treatment?
Dr. Kelly: I would check labs at baseline just to establish that the patient has normal thyroid levels. If they are low, then you’re dealing with a different disorder. You’d need to treat hypothyroidism first, and I would use T4 (levothyroxine, Synthroid) there, since that is the standard in the endocrinology practice guidelines (Garber JR et al, Endocr Pract 2012;18(6):988–1028). If the thyroid labs are normal and you are using thyroid as augmentation, then following the labs is not as relevant because there is no way to measure hypothyroidism at the cellular level. Generally, the thyroid levels are going to be high and the TSH low when you give high doses of thyroid, but not always. What matters is that the patient is not having symptoms of hyperthyroidism.

TCPR: What if the patient is already taking T4 to treat hypothyroidism?
Dr. Kelly: If the hypothyroidism is treated, I would consider adding T3 (triiodothyronine, Cytomel) for depression.

T3, T4, and Dosing

TCPR: When you are using thyroid to augment treatment in bipolar, do you use T3 (triiodothyronine, Cytomel) or T4 (levothyroxine, Synthroid)?
Dr. Kelly: Either works, and there’s no research telling us which is better. I prefer T3 because it’s easier to dose. T3 has a half-life of one day, while T4’s half-life is five to seven days, so you can get to steady state much faster with T3—in five days instead of four to five weeks. T3 is also more active in the brain, but this is not a big deal because about 40%–50% of T4 gets converted to T3 in the liver. Some people have trouble converting it, but that is very rare (about one in 5,000). Another difference is that T4 is not absorbed as well when patients eat foods that are high in calcium or fiber, while T3 has minimal food interactions.

TCPR: How do you dose T3?
Dr. Kelly: I start with 25 mcg for a week and then increase to 50 mcg daily. Then I see them back in two weeks and measure their heart rate and assess side effects and response. If they are not having side effects and have not recovered, then I will increase to 75 mcg for a week and then 100 mcg daily, and then see them back in two weeks. Sometimes I’ll only titrate by 12.5 mcg, such as in older patients or those who are worried about side effects. Most people respond somewhere between 50 and 150 mcg daily. When they do respond, it’s very quick. It’s like a light switch.

TCPR: Is it energy that improves, or the whole picture?
Dr. Kelly: It’s the whole picture. I see improvements in anxiety, sleep, and cognition, as well as depression.

TCPR: And how high do you go before you give up and say, “This patient is a nonresponder?”
Dr. Kelly: If there is no response to T3 150 mcg daily for two to four weeks, I would give up. If they are on carbamazepine, I may need to go higher, like 200–300 mcg, because carbamazepine lowers thyroid levels.

TCPR: How long do you continue T3 after a patient recovers?
Dr. Kelly: It’s hard to know. Textbooks say you should try to stop it after they’ve recovered for six months, but that’s just because we don’t have good controlled studies that go longer than that. In my experience, most people relapse if you try to lower the dose. Occasionally people do well when it is taken away, but most need it long term. Most of my patients on thyroid augmentation have been on it for three to 10 years.

TCPR: Thank you for your time, Dr. Kelly.