Printable Version

Home > Abstracts of Current Literature > Effect of partial T3 substitution for T4

Partial Substitution of Thyroxine (T4) with Tri-Iodothyronine in Patients on T4 Replacement Therapy: Results of a Large Community-Based Randomized Controlled Trial

Background: Thyroxine (T4) replacement is a common treatment for hypothyroidism. However, T4 replacement to achieve normal TSH levels may result in low free T3 levels. Anecdotal reports show patients complain of impaired psychological well-being and hypothyroid symptoms even though they were maintained on T4 replacement with normal TSH.

Objective: The study objective was to see the effects on psychological well-being and symptoms suggestive of hypothyroidism, when T4 treatment is supplemented with T3.

Methods: This study was a large, double-blind, randomized, cross-sectional, placebo controlled trial. Patients were included if they were between the ages of 18-75 years, T4 dose ≥100 mcg/day, TSH level recorded in last 15 months and the last level known to be within the local laboratory reference range, and the absence of T4 adjustments in the last three months. Exclusion criteria included a history of myocardial infarction, unstable angina or heart failure in the previous three months, thyroid cancer or secondary hypothyroidism, cholestyramine use, use of antidepressants in the previous three months, or amiodarone usage in the previous 12 months. Six hundred and ninety-seven patients were randomized to receive 1) 10 mcg T3 or matched tablet of 50 mcg T4 indistinguishable from T3 and 2) the remaining T4 dose which was the original dose minus 50 mcg. The study duration was 12 months with check-in evaluations for outcome measures at three and 12 months. An active placebo, T4, was used as the control group. The primary outcome measures were the general health questionnaire (GHQ) Likert scores and the GHQ categorical scores. The secondary outcome measures included: free T3 and T4, TSH (mean and median), sex hormone-binding globulin (SHBG), cholesterol, alkaline phosphatase, creatinine kinase, calcium, pulse rate, systolic BP, diastolic BP, weight, BMI, fat content, thyroid symptom questionnaire (TSQ), Hospital anxiety and depression questionnaire (HADS), neuromuscular (% no symptoms), and well-being (% better). The study had an 80% power to detect a 0.7 point difference between groups on the GHQ Likert scale and an intent-to-treat analysis was used.

Results: At the three-month check-in visit 37 subjects were lost to follow-up. At the 12-month check-in 47 subjects were lost to follow-up. Primary outcome measure results at three months for the GHQ Likert scores showed improvement for the placebo and T3 group (p<0.001) as compared to baseline results. There was a 0.47-point difference in GHQ scores for the comparisons between the treatment groups (95% CI, -0.26 + 1.12, p=0.218). This 0.47 difference is smaller than the point difference the authors used to determine the power of the study. A greater reduction in psychiatric caseness was seen in the T3 group (19.2%) as compared to T4 (26.6%) alone (odds ratio: 0.61, 95% CI 0.42-0.90, p=0.01). At 12 months, the GHQ categorical scores in the T3 group had risen (worsened, p=0.0034) and there were no differences between treatment groups (p=0.24). At three months, the mean free T4 value in the intervention treatment group had fallen from the upper part of the reference range to the lower part, which is significantly lower than the placebo group. At 12 months, the free T3 to T4 ratio in both groups fell significantly (p<0.001). SHBG values decreased from baseline and then increased at 12 months. Physical or biochemical measures showed no significant differences other than a slightly lower diastolic BP in the T4 group. There were no differences seen in TSQ scores, sleep, neuromuscular symptoms, HADS depression category, or visual analog scales, and the percentage of patients reporting that they felt better on direct questioning. The authors concluded that the study did not provide convincing evidence that there is a benefit of partial substitution of T4 by T3 in patients’ psychological well-being or symptoms suggestive of hypothyroidism.

Strengths: The results did not appear to be biased by the authors. The design of the study included important aspects, including randomization and placebo-control.

Weaknesses: The study did not finish with a sufficient amount of subjects to provide the calculated 80% power because of the large number of drop-outs, thus allowing the possibility of Type II error. The adverse reactions of the treatment groups were not given in the study. Multiple psychological scales were used which may have caused an increasing possibility of positive findings by chance in one or more of the measures. The fixed treatment regimen may not have provided a sufficient amount of T4 for individuals. There was a relatively large placebo effect seen in the study results.

Conclusion: The results of the study did not sufficiently prove the benefit of supplementing T4 doses with T3. More studies need to be done since the treatment for hypothyroidism is a life-long treatment. Future studies could be designed to allow patients to stay on their current therapeutic regimen of T4, and supplementing with low doses of T3 and titrating the dose to therapeutic response, as compared to decreasing the T4 dosage and supplementing with T3 doses in a fixed manner.

Saravanan, Ponnusamy, et.al. Partial substitution of thyroxine (T4) with tri-iodothyronine in patients on T4 replacement therapy: results of a large community-based randomized controlled trial. J Clin Endocrinol Metab. 2005; 90: 805-812.

Kelly Ellis, Pharm.D. Candidate