Discussion
In this retrospective cohort study, we demonstrated that the risk of developing hypothyroxinemia in patients who used mirtazapine was 64% higher than patients who did not use, after adjusting for important confounders.
There are several strengths in this study. To the best of our knowledge, this is the first study to report an association between mirtazapine and increased risk of hypothyroxinemia. The large sample size of our study helped to generate stable estimates. The use of electronic inpatients records provided complete measurements of thyroid function: T4, free T4, and TSH at baseline and during treatment. As a result, we could have complete ascertainment of hypothyroxinemia. Furthermore, we excluded patients who had thyroid dysfunction at baseline or who received mood stabilizers (i.e. lithium, valproic acid, and oxcarbazepine) or quetiapine during hospitalization and adjusted for a number of important confounding factors including the baseline level of free T4 and TSH and the use of other kinds of antidepressants during the treatment. Thus, the impact of prior thyroid dysfunction or other psychiatric medications on the observed association between mirtazapine use and hypothyroxinemia would have been minimized.
Limitations of this study should be acknowledged. First, we had no dynamic monitoring information of thyroid function. Although we have collected data on serum thyroid hormone levels at baseline and during hospitalization for antidepressants treatment, no additional follow up tests were performed. As a result, we could not determine whether the effect of mirtazapine on thyroid function persists after discharge from hospital.
Second, the clinical symptoms/signs of thyroid dysfunction were not available in this study. The laboratory criteria of hypothyroxinemia, defined as “normal TSH and Low free T4”, overlaps, to a great extent, with the laboratory criteria of central hypothyroidism3-5. Based on available data, we could not definitely differentiate hypothyroxinemia from central hypothyroidism. We adopted the term “hypothyroxinemia” for the “lower free T4 (<12.00 pmol/L) and normal TSH level” in our study.
Case reports 17-21 and clinical studies 22-26suggested that antidepressants other than mirtazapine may also be associated with thyroid abnormalities. However, no conclusion could be drawn from these studies due to small sample size23-26, cross-sectional study design 22, incomplete laboratory tests20,21,26. and lack information whether hormone concentrations reached cut off values23,24,26. On the other hand, our study found that the incidence of hypothyroxinemia in patients treated by SSRIs, SNRIs, or combined treatment in non-mirtazapine group was similar (data available upon request). Our findings, in principle, were consistent with previous studies/reviews that SSRIs and SNRIs reduced T4/FT4 levels but did not significantly affect TSH level27.
Compared with other treatments for major depressive disorder, the impact of mirtazapine treatment on thyroid function has not been well investigated. To date, only one study reported a significant decrease in free T4 in 17 outpatients affected by major depressive disorder after mirtazapine treatment for 6 months10. No change in TSH was observed and no comparison with non-mirtazapine users was made in that study 10. Although that study did not report if free T4 was dropped to the cutoff level defining hypothyroxinemia, the significant decrease in free T4 after mirtazapine treatment10 provides some support to our finding on mirtazapine and hypothyroxinemia association. That previous study also reported a significant increase in FT3 after mirtazapine treatment10, which was not observed in our study. On the other hand, our study observed that the degree of reduction in T4 and free T4 after mirtazapine treatment was much higher than the reductions in T3 and FT3, which suggests that peripheral conversion of T4 to T3 could be affected by mirtazapine treatment, and this process may occur mostly in liver, heart, muscle and gut, rather than in thyroid6. Increased peripheral conversion of T4 to T3 may also be a consequence of a compensatory response to lower function of hypothalamic–pituitary–thyroid axis, due to changes in thyroid hormone metabolism that serve to maintain normal circulating levels of T3 until the late stages of dysfunction6.
More than 99% of circulating thyroid hormones is bound to serum proteins. It is reported that commercially available testing kits might show inaccurately low free thyroxine concentrations in patients treated with drugs which have the capacity to displace thyroid hormone from binding proteins, and it is suggested that serum dilution in assays could reduce free thyroxine readout6. However, we observed that the degree of reduction was very similar in both T4 and free T4. For example, the reduction was 18.77% in T4 and 21.09% in free T4 in the mirtazapine group. It is known that serum dilution in assays would not affect T4 readout 6. Therefore, in this case, the low serum free T4 levels observed in patients treated with mirtazapine may not be due to artifactual laboratory effects.
The key to maintain the level of serum free thyroxine is the negative feedback regulation, an appropriate TSH synthesis and release in response to serum thyroid hormone level changes6. Activation of the central neurotransmitters system, such as dopaminergic and serotonergic system, has been reported to inhibit hypothalamic–pituitary–thyroid axis3,5,7. The increased risk of hypothyroxinemia after mirtazapine treatment could be explained by the central effect of mirtazapine on hypothalamic/pituitary prevents the compensatory increase in TSH in response to low T4, thereby ultimately presenting as higher incidence of hypothyroxinemia in patients treated by mirtazapine. It has been reported that antidepressants (other than mirtazapine), which improve central dopaminergic and serotonergic system, seem to decrease T4/free T4 levels but maintaining TSH level27. Our study found that after adjusting for other antidepressant treatments, the association between mirtazapine treatment and hypothyroxinemia remained significant.
The finding of increased risk of developing hypothyroxinemia after mirtazapine treatment may have significant clinical implication.13 Symptoms of thyroid dysfunction and major depressive disorder overlap in many aspects such as depressive mood, fatigue and weakness11. Thus, hypothyroxinemia could be easily ignored in major depressive disorder treated with mirtazapine, if attention is paid only to those with elevated TSH.
In addition, clinical studies indicate that thyroid hormone supplementation can improve the response rate and can also convert patients with refractory depression to responders15. The beneficial effects of thyroid hormone supplementation has suggested that a central hypothyroidism may exist in such patients15. Failure to recognize hypothyroxinemia or central hypothyroidism may result in ignored thyroid function insufficiency, which may lead to poor response to treatment in individuals with major depression disorder. It may therefore be important to identify thyroid function insufficiency during antidepressants treatment for major depressive disorder, especially in patients receiving mirtazapine as an adjunctive antidepressant.