| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Endocrinology and Diabetes (J.P.W., L.C.W., B.G.A.S., D.H., M.J.G.) and Psychiatry and Behavioural Science (L.S.), Sir Charles Gairdner Hospital and School of Medicine and Pharmacology, University of Western Australia (J.P.W., V.B., B.G.A.S.) and PathWest Laboratory Medicine WA (R.G., C.I.B., M.T.), Nedlands, Western Australia 6009
Address all correspondence and requests for reprints to: Dr. John P. Walsh, Department of Endocrinology and Diabetes, Sir Charles Gairdner Hospital, Nedlands, Western Australia 6009. E-mail: john.walsh{at}health.wa.gov.au.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to test whether adjustment of T4 dosage aiming for a serum TSH concentration less than 2 mU/liter improves well-being compared with a serum TSH concentration in the upper reference range.
Design: We conducted a double-blind, randomized clinical trial with a crossover design.
Participants: Fifty-six subjects (52 females) with primary hypothyroidism taking T4 (
100 µg/d) with baseline serum TSH 0.1–4.8 mU/liter participated.
Interventions: Each subject received three T4 doses (low, middle, and high in 25-µg increments) in random order.
Outcome Measures: Outcome measures included visual analog scales assessing well-being (the primary endpoint) and hypothyroid symptoms, quality of life instruments (General Health Questionnaire 28, Short Form 36, and Thyroid Symptom Questionnaire), cognitive function tests, and treatment preference.
Results: Mean (± SEM) serum TSH concentrations were 2.8 ± 0.4, 1.0 ± 0.2, and 0.3 ± 0.1 mU/liter for the three treatments. There were no significant treatment effects on any of the instruments assessing well-being, symptoms, quality of life, or cognitive function and no significant treatment preference.
Conclusions: Small changes in T4 dosage do not produce measurable changes in hypothyroid symptoms, well-being, or quality of life, despite the expected changes in serum TSH and markers of thyroid hormone action. These data do not support the suggestion that the target TSH range for the treatment of primary hypothyroidism should differ from the general laboratory range.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Anecdotal evidence suggests that some patients with primary hypothyroidism have improved well-being if the T4 dosage is titrated until serum TSH is in the lower part of the reference range (e.g. <2 mU/liter), and many authorities now recommend that this be the usual target for the treatment of hypothyroidism (5, 6, 7, 8). This approach has never been tested in a clinical trial, however, and there is no good evidence that the treatment target for primary hypothyroidism should in fact differ from the general laboratory reference range. A survey of U.S. endocrinologists and primary care providers published in 2001 found no consensus either within or between these physician groups as to what the target TSH range should be for T4 treatment (9). Although there is some controversy as to whether the upper limit of the TSH reference is inappropriately high (10, 11), an upper limit between 4 and 5 mU/liter remains widely used.
Only one previous study has examined the symptomatic effects of small adjustments of T4 dosage in patients with treated hypothyroidism (12). In this study, participants reported improved well-being when T4 was given at doses that suppressed TSH concentrations. The study was open label and not randomized, and its results might be explained by a placebo effect.
We therefore conducted a double-blind, randomized clinical trial with the aim of determining whether adjustment of T4 dosage aiming for a target serum TSH in the lower part of the reference range or below resulted in improved well-being and reduced symptoms of ill health compared with a target TSH in the upper part of the reference range.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Recruitment to the study commenced in May 2003, and the study was completed in March 2005. Inclusion criteria were treated primary hypothyroidism of at least 6 months duration, a stable T4 dose of at least 100 µg/d, no change in T4 dosage in the previous 2 months, and a serum TSH concentration of 0.1–4.8 mU/liter (measured at a screening visit). The diagnosis of hypothyroidism was confirmed from medical or laboratory records or by contacting primary care physicians; in a few patients with longstanding hypothyroidism, this was not possible. Exclusion criteria included major comorbidity, hypopituitarism, history of thyroid cancer requiring suppression of TSH secretion, cardiac disease, liothyronine treatment, and drug treatment known to affect thyroid hormone secretion, metabolism, or bioavailability or measures of thyroid hormone action.
Study design, treatments, and evaluation
The study had a double-blind, crossover design. Each subject received three T4 doses in turn, designated low dose, middle dose (25 µg more than low dose), and high dose (25 µg more than middle dose). The T4 doses were intended to result in serum TSH concentrations of 2.0–4.8 mU/liter during low-dose treatment, 0.3–1.99 mU/liter during middle-dose treatment, and less than 0.3 mU/liter during high-dose treatment, based on a previous study in which T4 dosage was adjusted by 25-µg increments and decrements (13). The order of treatment was randomized in permuted blocks of six, using sealed envelopes.
T4 dosage was adjusted using two methods, depending on the serum TSH concentration at the screening visit. Subjects with a serum TSH concentration of 2.0–4.8 mU/liter at baseline continued their usual T4 dose during the study and in addition took one capsule daily that contained placebo, T4 25 µg, or T4 50 µg (in random order). For these subjects, therefore, low-dose treatment consisted of their usual T4 dose, middle dose consisted of usual dose plus 25 µg daily, and high dose was usual dose plus 50 µg daily. Subjects whose serum TSH concentration was 0.1–1.99 mU/liter at baseline reduced their daily dose of T4 by 25 µg and took one capsule daily containing placebo, T4 25 µg, or T4 50 µg (in random order) in addition to their reduced T4 dosage. For these subjects, therefore, low-dose treatment consisted of their usual dose less 25 µg/d, middle dose was their usual T4 dose, and high dose their usual dose plus 25 µg/d. Treatment periods lasted 8 wk (without washout periods); in a few cases, the treatment periods were extended by one or more weeks to allow convenient scheduling of visits. T4 sodium was purchased from Sigma Pharmaceuticals Pty, Ltd. (Croydon, Victoria, Australia). (All T4 marketed in Australia is made by the same manufacturer, so issues of bioequivalence did not arise.)
At baseline and at the end of each treatment period, subjects attended after an overnight fast and before taking T4 or study medication. Assessments were carried out as described for a previous study (14). Venous blood and a random urine sample were collected for measurement of serum TSH, free T4, and free T3; SHBG and plasma cholesterol (markers of thyroid hormone action on liver) and plasma alkaline phosphatase and urine deoxypyridinoline (DPD)/creatinine ratio (markers of thyroid hormone action on bone). Symptoms and signs of hypothyroidism were assessed using the Billewicz scale as modified by Zulewski et al. (15), which gives a tissue hypothyroidism score out of 13. Resting pulse rate and blood pressure were measured in the supine position as cardiovascular markers of thyroid hormone action and ankle reflex relaxation time were assessed using a photomotogram (16). Treatment compliance was assessed by counting unused capsules.
At each visit, subjects were asked to complete 10 visual analog scales with regard to their symptomatic well-being over the previous few weeks. The scales assessed general well-being, happiness/sadness, confusion, anxiety, irritability, tiredness, feeling hot/cold, sickness/nausea, blurred vision, and aches/pains. Subjective satisfaction or dissatisfaction with each treatment was rated on a four-point Likert scale (14), and treatment preference was recorded at the final visit. Subjects self-administered three questionnaires: the Short Form 36 (SF-36) (17) as a generic quality of life instrument, the General Health Questionnaire 28 (GHQ-28) (18) as a measure of psychological function and disturbance, and the Thyroid Symptom Questionnaire (TSQ) (3) as a disease-specific instrument. The rationale for using these instruments and the visual analog scales has been discussed previously (14). The SF-36 was scored by standard methods (17). The GHQ-28 and TSQ were scored using a four-point Likert scale with each response scored 0, 1, 2, or 3 (3, 18). Cognitive function was assessed by a clinical psychologist using standard tests as previously described (14): the Symbol Digit Modalities Test (19), which assesses cognitive efficiency and ability to undertake a novel task; the Trail Making Test Parts A and B (20), assessing visual search, attention, mental flexibility, and motor function; and the Digit Span Sub-test (both Forward and Backward) of the Wechsler Adult Intelligence Scale III (21), assessing immediate auditory memory, attention, and concentration.
The primary outcome measure was the visual analog scale assessing general well-being. Based on the data of Carr et al. (12), we calculated that a sample size of 50 subjects would give 80% power to detect a 12-point difference between treatments, at 
= 0.05. We regarded a 12-point difference as clinically meaningful, because Carr et al. (12) found a 16-point difference on this visual analog scale between a biochemically optimal T4 dose and the dose preferred by patients. The recruitment target was set at 56, to allow for dropouts.
Biochemistry methods
Serum TSH, free T4, and free T3 were measured by chemiluminescence immunoassay on the Abbott Diagnostics Architect (Abbott Diagnostics, North Ryde, NSW, Australia). SHBG was measured by enzyme immunoassay using chemiluminescence substrate on Immulite 2000 (Diagnostic Products Corp., Los Angeles, CA). DPD was measured by an in-house ion-paired reversed-phase HPLC with fluorescence detection (22). Cholesterol and alkaline phosphatase were analyzed by standard biochemical methods on a Hitachi 917 analyzer (Roche Diagnostics, Indianapolis, IN). Intraassay and interassay coefficients of variation were as follows: TSH, 1.2 and 2.9%; free T4, 3.8 and 3.6%; free T3, 3.0 and 5.1%; SHBG, 4.1 and 6.0%, respectively; interassay coefficient of variation for DPD was 8.0%.
Statistical analysis
Descriptive statistics were generated using SPSS 12.5 (SPSS, Chicago, IL). Analysis used mixed models (PROC Mixed, SAS 9.1; SAS Institute, Cary, NC) with models that included terms for treatment, period, and sequence. Repeated measurements were modeled by random effects specification for each subject. Carryover effects were examined using treatment x period interactions. The primary analysis was an intention-to-treat analysis of the three treatments (low dose, middle dose, and high dose). The secondary analysis, prespecified in the protocol, was of data grouped by ranges of serum TSH concentration (2.0–4.8, 0.3–1.99, and <0.3 mU/liter) as measured at the end of the treatment periods. This was included because we anticipated that not all subjects would reach the target TSH concentrations after T4 dosage adjustment. Findings were considered significant at the 95% level.
Ethical approval
The study protocol was approved by the Human Research Ethics Committee of Sir Charles Gairdner Hospital. Informed consent was obtained from all participants. The study was registered with ClinicalTrials.gov (identifier NCT00111735).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
A total of 133 subjects were screened for the study (Fig. 1
), the majority following newspaper advertisement (108 subjects) and the remainder from endocrinology outpatient clinics (14 subjects) and participants in a previous clinical trial (11 subjects) (14). Of 26 eligible subjects who declined to enter the study, seven did so because they were unwilling to have their T4 dose reduced. Fifty-six subjects were recruited; their baseline characteristics are shown in Table 1. The mean (± SD) T4 dose at baseline was 120 ± 27 µg/d, equivalent to 1.7 ± 0.4 µg/kg·d; 33 subjects were taking T4 100 µg daily, and the remainder were taking up to 200 µg daily. The baseline serum TSH concentration was between 2.0 and 4.8 mU/liter in 17 subjects and between 0.1 and 1.99 mU/liter in 39 subjects.
|
|
The number of subjects randomized to each treatment sequence was as follows: low dose (L) followed by middle dose (M) followed by high dose (H), i.e. LMH, 10; LHM, nine; MLH, nine; MHL, nine; HLM, nine; and HML, 10 subjects. The mean (± SD) T4 doses during the study were 103.2 ± 4.1 µg/d (1.4 ± 0.1 µg/kg·d) for low dose, 127.2 ± 3.9 µg/d (1.8 ± 0.1 µg/kg·d) for middle dose, and 151.7 ± 3.9 µg/d (2.1 ± 0.1 µg/kg·d) for high dose.
Biochemical and clinical parameters
The mean serum TSH concentration was 2.8, 1.0, and 0.3 mU/liter in the low, middle, and high T4-dose groups, respectively (Table 2). As expected, mean free T4 and free T3 concentrations increased with increasing T4 dosage, and each was in the upper part of the reference range on high-dose treatment. A significant treatment effect was found on biochemical markers of thyroid hormone action, particularly serum SHBG and urine DPD/creatinine ratio. Plasma alkaline phosphatase activity showed less consistent effects, and plasma total cholesterol was significantly reduced during high-dose treatment compared with low dose. Fewer than expected subjects achieved TSH concentrations in the ranges 2.0–4.8 mU/liter and less than 0.3 mU/liter, leading to reduced numbers of subjects in the secondary analysis based on serum TSH at the end of the treatment periods (Table 2).
|
). There was no significant treatment effect on the Zulewski score of hypothyroid symptoms and signs or on pulse rate or blood pressure.
|
The results of the symptom scores and quality of life measures are shown in Table 4. For the primary endpoint of general well-being measured on a visual analog scale, there was no significant treatment effect, nor was there any significant difference between groups when analyzed by TSH concentrations achieved during treatment. No significant differences between groups were found on hypothyroid symptoms as measured by visual analog scales or the TSQ or on subjective satisfaction with thyroid replacement therapy. Quality of life as assessed by the summary scores or individual domains of the SF-36 instrument did not differ between treatment groups, nor did psychiatric health as measured by the overall GHQ-28 score or its subscales. In the secondary analysis by serum TSH concentrations achieved, the only significant difference was for the social dysfunction domain of the GHQ-28, which was significantly worse for the lower reference range (TSH, 0.3–1.99 mU/liter) than for the upper range (2.0–4.8 mU/liter). No significant treatment effect was detected on any of the tests of cognitive function (Table 5).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
There have been no previous randomized clinical trials addressing this question with which our data can be compared. Our results are, however, at variance with the common clinical impression that at least some hypothyroid patients feel well only if serum TSH concentrations are in the lower reference range or reduced. There are several possible explanations for this. First, individuals differ in their setpoints for thyroid hormone and TSH secretion and so presumably in tissue sensitivity to thyroid hormones (23, 24, 25). It is conceivable that there are symptomatic benefits of fine titration of T4 dosage in a subgroup of patients who were underrepresented in our study, either by chance or because such patients might be less likely to participate. Certainly, some screened subjects declined to participate because of the possibility that their T4 dose would be reduced, but the number of such subjects was small and the number of subjects screened and not randomized was not excessive. Second, it may be that our study was not powerful enough to detect a small, yet clinically relevant treatment effect, because our sample size calculations were based on an unblinded, nonrandomized study (12) rather than a randomized controlled trial. In that case, however, one would expect a trend toward benefit with increased T4 dosage, and none was evident. It is also conceivable that clinical benefit from T4 dosage change takes longer to manifest than biochemical steady state as reflected by thyroid function tests, and in this case, carryover effects could contribute to the negative findings of the study. However, the mean between-period difference in response to treatment based on the visual analog scale for well-being (the primary endpoint) was only 0.6 points, and differences related to treatment order were not consistent in sign. Although there was less than 30% power to detect a simple carryover effect of this order, it is clearly not clinically important and is unlikely to have prevented recognition of a significant treatment effect. Third, it may be that the instruments used to assess symptoms, quality of life, and cognitive function are insufficiently sensitive to detect small changes that are clinically relevant. Tentative support for this hypothesis can be drawn from two recent trials of combined T4/T3 treatment in which patients preferred combination treatment to T4 monotherapy despite a lack of measurable benefit on a range of quality of life and neurocognitive measures (26, 27). In our study, however, there was no evidence of patient preference for T4 doses that resulted in low-normal or suppressed TSH concentrations. Fourth, there is a strong placebo effect associated with thyroid hormone replacement (28), which may explain some or all of the apparent benefits of fine titration of T4 dosage in clinical practice.
Our results differ from the only previous clinical trial that has examined the effects of different T4 doses on well-being in treated hypothyroid subjects (12). In that study, participants reported improved well-being (measured on a visual analog scale) when treated with a dose of T4 50 µg/d greater than their optimal dose as determined by a TRH test. In most cases, serum TSH was suppressed to less than 0.2 mU/liter (the limit of assay sensitivity) on the increased T4 dosage. This discrepancy may be because we did not suppress TSH concentrations to as great an extent as did Carr et al. (12). However, it should also be noted that the study of Carr et al. (12) was open label and not randomized and that statistical analysis was by Student’s t test with no correction for multiple pairwise comparisons between groups, so its conclusions may not be secure.
Our results are perhaps not surprising when put in the context of randomized controlled trials of T4 treatment for subclinical hypothyroidism. Of five such studies, only one found convincing benefits of treatment (29), whereas in the remainder, benefits were marginal or unconvincing (30, 31) or not significant compared with placebo (32, 33). Because no significant symptomatic benefit over placebo was observed with T4 treatment that resulted in a fall of mean TSH from 11.7 to 3.1 mU/liter (32) or from 8.0 to 3.4 mU/liter (33), it is perhaps unreasonable to expect such a benefit associated with lesser reductions in serum TSH as in the current study.
The strengths of our study include its large sample size and crossover design, which is more powerful than a parallel design. Compliance with study medication was good, and the dropout rate was not excessive. A weakness in the study design was the use of fixed 25-µg adjustments in T4 dosage regardless of T4 dose at baseline and body weight, and it may be partly because of this that not all subjects achieved TSH concentrations in the target ranges of 2.0–4.8, 0.3–1.99, and less than 0.3 mU/liter. Ideally, serum TSH would have been checked during each treatment period and additional dosage adjustments made as necessary. We elected not to do this because of the potentially adverse effects of extra study visits on recruitment and retention and the difficulties inherent in adjusting dosage while preserving blinding. Most of the study participants were female (reflecting the female preponderance in this disorder), and it cannot be assumed that the results apply to males.
In conclusion, we found that in subjects with primary hypothyroidism, adjustment of T4 dosage with the aim of achieving serum TSH concentrations in the lower reference range or slightly below did not result in measurable symptomatic benefit compared with target TSH concentrations in the upper reference range. Our data do not support the suggestion that the TSH target range for the treatment of hypothyroidism should differ from the general laboratory reference range.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosures: The authors have nothing to declare.
First Published Online May 2, 2006
Abbreviations: DPD, Deoxypyridinoline; GHQ-28, General Health Questionnaire 28; SF-36, Short Form 36; TSQ, Thyroid Symptom Questionnaire.
Received January 17, 2006.
Accepted April 21, 2006.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  W. M Wiersinga Do we need still more trials on T4 and T3 combination therapy in hypothyroidism? Eur. J. Endocrinol., December 1, 2009; 161(6): 955 - 959. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Panicker, P. Saravanan, B. Vaidya, J. Evans, A. T. Hattersley, T. M. Frayling, and C. M. Dayan Common Variation in the DIO2 Gene Predicts Baseline Psychological Well-Being and Response to Combination Thyroxine Plus Triiodothyronine Therapy in Hypothyroid Patients J. Clin. Endocrinol. Metab., May 1, 2009; 94(5): 1623 - 1629. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Vaidya and S. H S Pearce Management of hypothyroidism in adults BMJ, July 28, 2008; 337(jul28_1): a801 - a801. [Full Text]
|
|
|
|
|
|
M. H. Samuels, K. G. Schuff, N. E. Carlson, P. Carello, and J. S. Janowsky Health Status, Mood, and Cognition in Experimentally Induced Subclinical Thyrotoxicosis J. Clin. Endocrinol. Metab., May 1, 2008; 93(5): 1730 - 1736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Schlechte Update in Endocrinology Ann Intern Med, October 16, 2007; 147(8): 563 - 572. [Full Text] [PDF]
|
|
|
|
|
|
M. H. Samuels, K. G. Schuff, N. E. Carlson, P. Carello, and J. S. Janowsky Health Status, Mood, and Cognition in Experimentally Induced Subclinical Hypothyroidism J. Clin. Endocrinol. Metab., July 1, 2007; 92(7): 2545 - 2551. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 What Thyroxine Dose Is Optimal for Patients with Hypothyroidism? Journal Watch (General), August 3, 2006; 2006(803): 1 - 1. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
