Table 3. Description and Results of Randomized Trials of Thyroxine Replacement Therapy^a

Table 3a. Known History of Thyroid Disease
Table 3b. No Known History or Not Stated
Table 3c. Biochemically Euthyroid Patients

Table 3a. Known History of Thyroid Disease

Study, Year	Patients	Setting	Age, Gender	Eligibility Criteria	Other Population Characteristics	Exclusion Criteria
Cooper et al., 198459 (more details)	Previously treated Graves' disease, stage C subclinical hypothyroidism	Thyroid specialty clinic, Boston, MA	32 women and 1 man; mean age, 55 yrs	TSH > 3.5 mU/L on 2 occasions	History of Graves' disease	None stated
Meier et al., 200176 (more details)	Autoimmune thyroiditis (n = 33), previously treated Graves' disease (n = 22), previously treated goiter (n = 7)	Thyroid specialty clinic, Switzerland	63 women; mean age, 58.5 ± 1.3 yrs	Women 18-75 yrs; TSH >6.0 mU/L on 2 occasions; exaggerated TSH response to TRH; good health	History of autoimmune thyroiditis (n = 33), Graves' disease (n = 22), goiter (n = 7).  Only 4 had idiopathic subclinical hypothyroidism	Coronary heart disease, lipid-lowering drugs, history of poor compliance  (estrogen therapy allowed)
Caraccio et al., 200278 (more details)	Hashimoto's thyroiditis (n = 48) or Graves' disease (n = 1)	Medical school internal medicine clinic, Italy	42 premenopausal women, 7 men	TSH > 3.6 mU/L for > 6 mos,  + atP and anti-Tg, good general health	Subjects had higher TC, LDL, and ApoB levels than healthy controls	Diabetes, renal or liver disease, TC > 7.8 mmol/L

Study, Year	Funding Sources and Role of Funder	Interventions (Dose, Duration)	Control	Baseline TSH Level	Number Screened/ Eligible/ Enrolled	Patients Withdrawn/ Analyzed	Outcomes Assessed/ When Assessed
Cooper et al., 198459 (more details)	U.S. Public Health Service (armour supplied LT4)	LT4, 50 mg, then titrated up	Placebo	11 (mean)            3.6-55.3 (range) mean TSH in control group increased to ~15 by the end of the study	656/91/41	8/33	Symptoms, lipid profile at 1 yr
Meier et al., 200176 (more details)	Swiss Research Foundation, Henning Berlin, Sandoz, Roche	LT4 titrated over 6 mo (mean final dose, 85.5 ± 4.3), with similar visits and changes in control group.  Total followup, 50 wks	Placebo	12.8 (mean) 5-50 (range)	NR/NR/66	3/63	Symptoms, lipid profile at 1 yr
Caraccio et al., 200278 (more details)	Grant from university	LT4, 25 then titrated up	Placebo	5.43 (mean) 3.65-15 (range)	NR/NR/49	0/49	Lipid profile at 6 mos for placebo group vs about 11 mos for LT4 group

Study, Year	How Were Symptoms Assessed (e.g., Scales Used)?	LT4 vs Placebo Group Results	Before/After Results	Adverse Effects Assessed?	Adverse Effects	Quality Rating (Good/ Fair/ Poor)	Relevance to Screening	Comments and Questions
Cooper et al., 198459	Symptom change scores (Cooper Questionnaire)	Improved symptoms (-1.2 vs 2.1) in LT4 group.  47% improved in LT4 group vs 19% in placebo group (NNT = 3.6).   No difference in lipid profiles	Placebo group's TSH and symptoms rose during the year, suggesting the patients had rapidly advancing subclinical hypothyroidism	Only through symptom scores	4 patients in LT4 group felt worse, vs 6 in placebo group	Good	Low	Well-conducted trial, but subjects had known thyroid disease and the study is not relevant to screening.  What proportion of all patients who had elevated TSH and normal FT4 were eligible for the study?
Meier et al., 200176	Thyroid symptom questionnaire	Post-treatment LDLc was the same in both groups (3.7 ± 0.2, P = 0.11), and symptoms scores were not significantly different (P > 0.2)	LDLc reduced from 4.0 to 3.7 in the LT4 group (P = 0.004) and there were borderline improvements in symptom scores (P = 0.02).  Placebo group TSH was stable	No	Not assessed	Poor	Low	The discrepancy between before/after results and LT4 vs placebo results suggests that randomization was probably flawed. Were patients informed of their LDLc levels?
Caraccio et al., 200278	Not assessed	No significant differences between LT4 and placebo groups in any lipid variable	LT4 group:  TC reduced from 5.5 to 5.0; LDLc from 3.6 to 3.1	No	Not assessed	Poor	Fair	Analyzed as an open, uncontrolled study. Was completion of the study a criterion for inclusion in the analysis?  How many patients were screened, eligible, enrolled, and randomized?  Were patients & providers aware of treatment?  How was randomization done?  Were baseline differences statistically significant?  What proportion of subjects in each group had a TC > 6.2?

Top of Page

Table 3b. No Known History or Not Stated

Table 3a. Known History of Thyroid Disease
Table 3b. No Known History or Not Stated
Table 3c. Biochemically Euthyroid Patients

Study, Year	Patients	Setting	Age, Gender	Eligibility Criteria	Other Population Characteristics	Exclusion Criteria
Jaeschke et al., 199674 (more details)	Diagnosis of subclinical hypothyroidism	Unclear setting, Ontario, Canada	28 women and 9 men over age 55; mean age, 68 yrs	TSH > 6 mU/L on 2 occasions	None stated	Medications that interfere with thyroid function test results; serious medical conditions
Kong et al, 200279 (more details)	Women with a diagnosis of subclinical hypothyroidism	Referrals from GPs for thyroid function tests, London, UK	45 women; mean age, ~49 yrs	Women over 18 yrs; TSH level, 5 < 10 mU/L	Most patients were referred because of symptoms	History of thyroid disease, psychiatric disorder, anticipated pregnancy
Nystrom et al., 198858 (more details)	Women identified by screening	Population-based screening study, Gothenburg, Sweden	20 women; aged 51-73	Women over 18 yrs; TSH level, 4< 15 mU/L, exaggerated TSH response to TRH	Symptoms did not differ between subjects and healthy controls	History of or signs of thyroid disease, history of cardiovascular disease

Study, Year	Funding Sources and Role of Funder	Interventions (Dose, Duration)	Control	Baseline TSH Level	Number Screened/ Eligible/ Enrolled	Patients Withdrawn/ Analyzed	Outcomes Assessed/ When Assessed
Jaeschke et al., 199674 (more details)	Ontario Ministry of Health, Boots Pharmaceuticals	LT4 25 then titrated up (mean final dose, 68 ± 21)	Placebo	9.4 (mean) 6-32 (range)	NR/NR/37	6/31	Quality of life, symptoms, lipid profile at 6 mos
Kong et al, 200279 (more details)	Medical Research Council	LT4, 50 then titrated up to 100 if TSH >6 mU/L	Placebo	~7.7 (mean)	NR/52/45	10/34 (for quality of life) 18/27 for lipids	Quality of life, symptoms, lipid profile at 6 mos
Nystrom et al., 198858 (more details)	Non-industry grants (Nyegaard supplied LT4)	LT4, 50 for 2 wks, then 100 mg for 2 wks, then 150 daily	Placebo	~7.7 (mean) 2.9-16.3 (range)	1,192/22/20	3/17	Quality of life, psychometric symptoms, vital signs, ECG, lipid profile at 6 mos

Study, Year	How Were Symptoms Assessed (e.g., Scales Used)?	LT4 vs Placebo Group Results	Before/After Results	Adverse Effects Assessed?	Adverse Effects	Quality Rating (Good/ Fair/ Poor)	Relevance to Screening	Comments and Questions
Jaeschke et al., 199674	Chronic Thyroid Questionnaire, Cooper Questionnaire, SIP, cognitive tests	No improvement in symptoms or lipids; improved memory in LT4 group (mean difference of 0.58 on z-score scale, described as "small and of questionable clinical importance")	Placebo group's TSH rose from 9.42 to 10.32 over 6 mos	Only through dropouts	1 case of atrial fibrillation and 1 case of angina in LT4 group	Fair	Fair	Description of recruitment was inadequate. Were patients referred from family practitioners? Were patients who had a history of thyroid disease included?
Kong et al, 200279	Thyroid symptom questionnaire, GHQ-30, HADS	No improvement in symptoms or lipids	Placebo group's TSH dropped from 7.3 to 5.6 over 6 mos	Only through symptom scores	Anxiety scores were higher in the LT4 group	Poor	Fair	High dropout rate, but patients were relevant to primary care:  symptomatic with borderline TSH values
Nystrom et al., 198858	Thyroid symptom questionnaire, reaction time, Bingley's memory test	No difference in lipids	Symptom scores improved by the equivalent of 1 symptom per subject (P < 0.001), and 4 patients felt better with LT4 than with placebo	Only through dropouts	In LT4 group, 1 subject dropped out because of nervousness, 1 because of a sense of tachycardia	Poor	Good	The flaws in analyzing data make the study uninterpretable, but the patients are most like those encountered in screening

Top of Page

Table 3c. Biochemically Euthyroid Patients

Table 3a. Known History of Thyroid Disease
Table 3b. No Known History or Not Stated
Table 3c. Biochemically Euthyroid Patients

Study, Year	Patients	Setting	Age, Gender	Eligibility Criteria	Other Population Characteristics	Exclusion Criteria
Michalopoulou et al., 199875 (more details)	Patients referred for lipid assessment	Preventive medicine (lipid)  hospital-based clinic, Greece	Not stated	TC>7.5 mmol/L and TSH 0.4-4.0 mU/L	None stated	Conditions and medications that affect lipid profiles
Pollock et al., 200177 (more details)	Symptomatic patients with normal serum free thyroxine and TSH levels.	Referrals from GPs, hospital clinic, and response to newspaper ad, Glasgow, UK	25 symptomatic and 19 asymptomatic subjects, sex and age not given	Recent thyroid function tests within the reference range plus (a) at least 3 symptoms of hypothyroidism (tiredness, lethargy, weight gain, or 3 others) or (b) no symptoms	Symptomatic patients weighed more and had worse memory and psychological function than healthy controls	Current medical disorders

Study, Year	Funding Sources and Role of Funder	Interventions (Dose, Duration)	Control	Baseline TSH Level	Number Screened/ Eligible/ Enrolled	Patients Withdrawn/ Analyzed	Outcomes Assessed/ When Assessed
Michalopoulou et al., 199875 (more details)	Not stated	LT4, 50	LT4 25 mg	Stratified: 1.0 (mean) or ~2.6 (mean)	NR/NR/110	0/110	Lipid profile
Pollock et al., 200177 (more details)	Association of Clinical Biochemists	LT4, 100	Placebo	1.9 (mean)	NR/NR/25b	3/22b	Symptoms, vital signs, biochemical tests after 14 wks

Study, Year	How Were Symptoms Assessed (e.g., Scales Used)?	LT4 vs Placebo Group Results	Before/After Results	Adverse Effects Assessed?	Adverse Effects	Quality Rating (Good/ Fair/ Poor)	Relevance to Screening	Comments and Questions
Michalopoulou et al., 199875	Not assessed	LDL reduced from 6.2 to 6.1 in 25-mg group and from 6.8 to 5.9 in 50-mg group	LDLc reduction was significant in 50-mg group	No	Not assessed	Poor	Fair	Description of recruitment was inadequate. Were patients referred from family practitioners? Were patients who had a history of thyroid disease included?
Pollock et al., 200177	SF-36 plus validated cognitive/memory testing	Among symptomatic patients (n = 22), there were no important differences between LT4 and placebo groups in any SF-36,  memory, or cognitive measures	Placebo significantly improved SF-36 general health and physical health scores	Not assessed, except for SF-36 scores	In asymptomatic patients, LT4 significantly reduced SF-36 vitality scores	Fair	N/A	Too small; authors note that it is only a "pilot study."  Placebo effect, adverse effect of LT4 in healthy subjects, and baseline difference in cholesterol levels (6.3 vs 5.2) between symptomatic and asymptomatic subjects deserve more study

Top of Page

^a Values reported with a plus/minus sign are means ± SD.
^b Symptomatic Group Only.

Note: Anti-TG, anti-thyroglobulin; ApoB, Apolipoprotein B; atP, antithyroid-peroxidase; ECG, electrocardiogram; GHQ, General Health Questionnaire; GP, general practitioner; HADS, Hospital Anxiety and Depression Questionnaire; LDL, low-density lipoprotein; LDLc, low-density lipoprotein cholesterol; LT4, levothyroxine; NNT, number needed to treat for benefit; SF-36, Medical Outcomes Study Short Form; SIP, Sickness Impact Profile; TC,  total cholesterol; TRH, thyrotropin-releasing hormone; TSH, thyroid-stimulating hormone.

Return to Document