Reviewed by Dr. Laurette Femnou Mbuntum
Stott D, et al. N Engl J Med. 2017 Jun 29;376(26):2534-2544. doi: 10.1056/NEJMoa1603825

Subclinical hypothyroidism is defined as an elevated thyroid stimulating hormone (TSH) with a normal free thyroxine (T4) level. There is little evidence to support thyroid hormone replacement versus monitoring especially when patients have non-specific symptoms that might be attributed to hypothyroidism. This study is a randomized, double-blind, parallel-group trial analyzing the effect of normalizing the TSH level. Participants were 65 years of age or older with persistent subclinical hypothyroidism on at least two measurements, at least three months apart. Participants in the intervention group received levothyroxine starting at 50 mcg daily, and the dose was increased until TSH was within normal range. Control patients received a placebo with mock adjustment of the dose throughout the study. The primary end point was change in thyroid symptom related quality of life measured using ThyPRO and Tiredness scale scores. Secondary outcomes included several quality of life measurements as well as changes in weight, waist circumference, blood pressure, and body mass index. Patients were followed for one to three years. There were 737 total participants, 368 in the levothyroxine group and 369 in the control group. There was no significant difference in baseline characteristics or drop-out rate between the two groups. There was a statistically significant difference (p < 0.001) between the groups in average TSH level at both one and three year follow up: 3.63 mIU/L and 3.47 mIU/L in the treatment group, at one and three year follow up respectively, versus 5.48 mIU/L and 5.28 mIU/L in the placebo group (normal range 0.4-4.59 mIU/L). There was no significant difference in the primary outcome of hypothyroid symptoms and tiredness scores at one year. At three years, there was a clinically small magnitude 3.5 point decrease in tiredness score (p 0.05) in the levothyroxine group compared to the placebo group, but no difference in hypothyroid symptoms between the groups. One secondary outcome scale showed a small deterioration in quality of life with levothyroxine at one year (p 0.05) but a small improvement with levothyroxine at three years (p 0.03). Otherwise, there was no significant difference in secondary endpoints between the groups. There were 78 serious adverse events in the levothyroxine group compared to 103 adverse events in the placebo group (p 0.049).

Subclinical hypothyroidism is frequently encountered in older adults in both the hospital and clinic, and the clinical significance is unclear, as symptoms of hypothyroidism may be non-specific and can often be attributed to other chronic conditions. This study is one of the largest randomized controlled trials to evaluate the clinical effects of treating subclinical hypothyroidism and showed no appreciable difference in symptoms or the secondary outcomes of muscle strength, body mass index, blood pressure, and waist circumference. This reinforces the current recommendation of not treating subclinical hypothyroidism with TSH <10 mIU/L in older patients, despite nonspecific symptoms. One limitation of the study is that the average TSH was only 6.4 mIU/L, and it has been previously noted that older adults may have a slightly higher normal TSH than younger adults, the clinical significance of which is unknown. There were few patients with TSH higher than 10 mIU/L, which is the level at which treatment would be considered. In addition, thyroid antibody levels were not measured. Patients with positive antibodies are more likely to progress to overt hypothyroidism than those without antibodies. More studies are needed to determine the effect of thyroid replacement in subclinical hypothyroidism with TSH level higher than 10 mIU/L in older adults.