Current issues of ACP Journal Club are published in Annals of Internal Medicine


Prognosis of aubclinical thyrotoxicosis

ACP J Club. 1991 Jan-Feb;114:23. doi:10.7326/ACPJC-1991-114-1-023

Source Citation

Tenerz A, Forberg R, Jansson R. Is a more active attitude warranted in patients with subclinical thyrotoxicosis? J Intern Med. 1990;228:229-33.



To evaluate the clinical course of subclinical thyrotoxicosis.


An inception cohort was formed of patients with normal free thyroxine (FT4) but suppressed thyroid stimulating hormone (TSH) levels. A control group, matched for age and sex, was formed of persons with normal FT4 and TSH. Both groups were followed for 2 years.


A regional laboratory for thyroid function testing in Vasteras, Sweden.


By screening 5239 consecutive blood samples referred for thyroid tests, 112 untreated patients were identified who met criteria for study inclusion, and 40 were selected because they lived within the municipality of Vasteras. Patients with transient thyroiditis or depressive illness were excluded. 40 controls with normal thyroid function were chosen from the original referrals.

Assessment of prognostic factors

Subclinical thyrotoxicosis was detected by assay of serum samples and defined as TSH below 0.1 mU/L and normal free fractions of thyroid hormone.

Main outcome measure

Follow-up testing included ECGs and serum levels of TSH, FT3, and FT4. Diagnosis of nodular goiter was by blinded clinical examination. Thyroid scintigraphy or cytologic or histologic examination of surgical specimens was also sometimes used.

Main results

At baseline mean age was 65 years with a ratio of 7 women to each man in both groups. Reasons for referral for testing included suspected thyrotoxicosis (subclinical thyrotoxicosis [ST] group, 37%; controls, 42%), suspected hypothyroidism (8%; controls, 28%), goiter (35%; controls, 15%), or atrial fibrillation (15%; controls, 5%). At follow-up 2 years later, 12 (30%) of the ST patients and none of the controls had been treated for hyperthyroidism (P < 0.01). Atrial fibrillation occurred significantly more often in the ST group (11 patients [28%] compared with 4 [10%], P < 0.05). 27 patients in the ST group had nodular goiter compared with 8 in the control group. 6 patients in the ST group died compared with 5 in the control group. [The sample size to detect an important difference in mortality was not reported.]


In 2 years of follow-up, elderly patients with subclinical thyrotoxicosis were more likely to develop, and be treated for, clinical thyroid disease. They had a higher prevalence of nodular goiter and a greater likelihood of developing atrial fibrillation than euthyroid patients. ST had no apparent effect on mortality [but the sample size was too small for this result to be reliable].

Source of funding: Not stated.

Address for article reprint: Dr. A. Tenerz, Med klin, Centrallasarettet, S-721 89 Vasteras, Sweden.


The new “sensitive” TSH assay is able to distinguish truly normal from suppressed levels of TSH in patients with normal thyroid hormone levels. As many as 17% of patients with low TSH and normal thyroid hormone levels have nonthyroid illnesses (1). Others may have the very earliest sign of an autonomous thyroid gland, “subclinical thyrotoxicosis” (ST). This article validly assesses the yield of continued surveillance of patients with ST. Although the numbers were small, there were more new cases of atrial fibrillation among ST patients. However, only two patients developed unequivocal biochemical evidence of thyrotoxicosis. This 5% incidence is comparable to the 4% incidence found in another study from Sweden (2).

The free FT4 index may be more cost-effective than TSH in screening patients (3). Of course, by definition, ST could not be diagnosed by a free FT4 index alone. But, if most patients with biochemical ST do not have or develop overt thyroid disease, is it worthwhile to screen such patients?

Merville Marshall, MD
Long Island Jewish Medical Center New Hyde Park, New York, USA