This article is a collaboration between ֱ and:
A 75-year-old woman presents to your clinic for a chronic disease visit. She has atrial fibrillation, hypertension, and osteoporosis. Her medications include warfarin, metoprolol, lisinopril, hydrochlorathiazide, atorvastatin, alendronate, calcium/vitamin D, and levothyroxine. The levothyroxine was started several years ago when, on a routine lab check, her primary care doctor found her thyrotropin (TSH) to be 8 mIU/L, with a normal free T4 level. She reports chronic back pain following vertebral compression fractures last year. On exam, she appears more hunched over than usual, consistent with worsening vertebral osteoporosis. Her TSH level returns at 0.10 mIU/L (reference range: 0.4-4.0 mIU/L).
We suspect many clinicians would hardly find this case noteworthy. Many would lower her dose of levothyroxine and recheck the TSH in a few months.
We think, however, that this case highlights a problem of critical importance: the overuse of thyroid replacement, particularly among older patients. In our experience, patients with nonspecific symptoms like fatigue, low energy, or constipation receive a TSH check as part of the work up. If the level is elevated (or even on the high end of normal), the treating clinician will frequently start levothyroxine.
Yet levothyroxine has important side effects, and now published last month in the New England Journal of Medicine demonstrates that the treatment of mildly elevated TSH levels doesn't have benefit when the free thyroxine (free T4) level is normal (note that this situation is referred to as "subclinical hypothyroidism," which is a confusing misnomer since the definition is not based on whether or not the patient has symptoms attributable to their thyroid hormone status).
In the study, adults 65 years of age and older with elevated TSH levels (4.6-19.99 mIU/L) and a free thyroxine level within the reference range were randomized to receive either levothyroxine titrated based on the TSH level or placebo. After a year, levels of "tiredness" and scores on a "thyroid-related quality of life" survey were not different in the two groups.
Previous research has also shown that, like in the patient described above, when levothyroxine is initiated, the dose is often too high. found that an astounding 5.8% of patients started on thyroid replacement have an excessively low TSH (<0.1 mIU/L) after 5 years!
Moreover, overtreatment of hypothyroidism has clear downsides that frequently go unrecognized. Patients like the one in the vignette who have been on years of thyroid therapy for uncertain reasons may develop complications, such as atrial fibrillation or osteoporosis.
What's the take-home message? While asymptomatic patients with subclinical hypothyroidism (i.e., elevated TSH but normal free T4) require close monitoring, they frequently should not receive thyroid replacement unless their TSH levels are persistently and substantially above the reference range (perhaps even as high as 20 mIU/L). For the patient in the vignette, we would recommend not only lowering the dose but perhaps even considering a complete taper to assess whether the medication is necessary (with close monitoring to ensure she does not subsequently develop frank hypothyroidism).
Even trickier, however, are patients with elevated TSH levels who also have substantial non-specific symptoms that may be the result of a thyroid disorder, such as fatigue or depression. In most cases, these common non-specific symptoms are completely unrelated to thyroid hormones; however, patients with such symptoms were not the focus of the new NEJM study (though they were not excluded either). Ideally, a future study will assess the value of thyroid therapy in a population all of whom have non-specific symptoms as well as subclinical hypothyroidism.
In the meantime, our "Slow Medicine" principles suggest that we should be cautious about initiating thyroid therapy among such patients, because, in most cases, the symptoms are unlikely the result of thyroid disease. When we do initiate thyroid therapy, we should do so cautiously, at a low dose, and with close monitoring of both benefits and side effects, being vigilant to discontinue the medication if the patient's symptoms do not improve.
"Updates in Slow Medicine" applies the latest medical research to support a thoughtful approach to clinical care. It is produced by , of the Keck School of Medicine at the University of Southern California, and , of Harvard Medical School. To learn more, .