Thyroid augmentation  The addition of thyroid hormone to an antidepressant in antidepressant nonresponders has been reported since the late 1960s. In one case series of women, thyroid hormone, when prescribed in addition to an antidepressant, yielded a faster response than did an antidepressant alone (Prange et al. 1969; Wilson et al. 1970). Later case reports and series extended these results, suggesting that adding thyroid hormone to an antidepressant after a trial of that antidepressant had failed was beneficial for different types of patients (Banki 1975; Earle 1970; Ogura et al. 1974; Tsutsui et al. 1979). Triiodothyronine (T₃) (25 μg) given to 21 unipolar depressed patients who had not responded to unspecified doses of a variety of antidepressants for 4 weeks led to significant improvement (Targum et al. 1984). The highest reported success rate for thyroid hormone augmentation was 53% for T₃ compared with 19% for placebo (Joffe et al. 1993c).

In contrast, no difference was observed between T₃ 25 μg/day and placebo added to a failed 4-week trial of imipramine (mean dosage = 206 mg/day; mean imipramine plus desipramine blood levels = 200 ng/dL) in 16 unipolar depressed patients (Gitlin et al. 1987). No drug effect was detected because all patients improved in the 4 weeks following the imipramine trial.

In an open 4-week trial of T₃ at 25 μg daily in 20 outpatients who did not respond to 12 weeks of imipramine (mean dosage = 240 mg/day) plus IPT, the response rate was no different from that of a previous group that continued to receive imipramine without T₃ augmentation (Thase et al. 1989b).

In a 1990 survey of 118 psychiatrists asked to choose the next treatment in a vignette of an inpatient who had not responded to 4 weeks of treatment with nortriptyline, only 1 respondent (0.8%) chose to add T₃ as an augmentation strategy (Nierenberg and Cole 1991).

Case series and anecdotal reports, but not randomized controlled trials, have been reported for thyroid augmentation of SSRIs (Crowe et al. 1990; Gupta et al. 1991; Joffe 1992). A meta-analysis showed that, in the aggregate, thyroid augmentation is more effective than placebo (R. Aronson et al. 1996).

Adverse effects associated with thyroid augmentation are typically minimal but may include increased anxiety, jitteriness, tachycardia, insomnia, and sweating. In addition, thyroid hormones can increase atrial irritability and/or ventricular function (rate, force, and velocity of contractions) and can potentially lead to high-output failure. Caution should be exercised when giving thyroid hormone to patients with cardiac insufficiency or to elderly patients. Long-term treatment with high doses of thyroid hormone has also been associated with an increased risk of osteoporosis.

How much thyroid hormone should be given and for how long?  Most reports have involved dosing ranges of 25-50 μg of T₃, in the form of liothyronine (e.g., Cytomel), with lower doses in elderly patients and higher doses (e.g., 75 μg) in some patients. For thyroxine (T₄), usual thyroid hormone replacement therapy doses (e.g., 150 μg) have been used for the treatment of patients with refractory depression. Higher doses of both T₃ and T₄ may be considered for selected patients as long as side effects are monitored carefully.

To determine possible differences in efficacy between T₃ and T₄ in the augmentation of antidepressants, 37.5 μg of T₃ or 150 μg of T₄ was given for 3 weeks to 38 euthyroid unipolar depressed patients who had undergone failed adequate trials of either imipramine or desipramine (Joffe and Singer 1991). Fifty-three percent (9/17) responded to T₃, whereas only 19% (4 of 21) responded to T₄, a statistically significant difference. No guidelines exist for the duration of an adequate acute trial of thyroid hormone augmentation, but most patients who respond do so within 2-3 weeks. The longest reported duration of a thyroid augmentation study is 4 weeks (Thase et al. 1989b).

As with lithium augmentation, the literature is not a particularly helpful guide to clinical practice because few longitudinal studies and no double-blind discontinuation studies of responders to thyroid augmentation are available. A reasonable strategy is to test whether the adjunct remains necessary for maintaining a response by a trial of discontinuation 8-12 weeks after a response. If the patient experiences a relapse during the discontinuation trial, the thyroid hormone adjunct can be reinstated.

What are the predictors of response to thyroid augmentation?  No significant predictors of response have been found in the only study to search for them (Joffe et al. 1993b).

Conclusion  The evidence for thyroid augmentation of antidepressants for euthyroid unipolar depressed patients suggests that some patients may benefit from this strategy. In one controlled study it was found that thyroid augmentation was superior to placebo and equal to lithium augmentation (Joffe et al. 1993c). Studies to clarify the duration of an adequate acute trial and the utility of maintenance therapy with this adjunct are needed.