Levothyroxine

Levothyroxine replaces a missing hormone. For people with overt hypothyroidism, the evidence that it prevents myxedema coma and reverses severe symptoms is strong. But for the much larger group diagnosed with "subclinical" hypothyroidism, the picture is different.

Two large, well-done randomized trials in older adults with subclinical hypothyroidism found that levothyroxine did not improve symptoms or tiredness compared to placebo [1, 2]. The trials were placebo-controlled, not funded by the manufacturer in a way that would bias the primary result, and they measured real outcomes (how people felt), not just a lab number. In the 2017 NEJM trial, after a year, the levothyroxine group's TSH dropped from about 6.4 to 3.6 mIU/L, but their hypothyroid symptom scores and tiredness scores changed by the same trivial amount as the placebo group [2]. The 2017 Annals trial of 737 patients found the same: zero benefit on symptoms [1].

This is the key finding the standard narrative skips. Lowering a lab number is not the same as helping a person. The intervention is given to fix a symptom, and when tested against placebo in the population where it is most commonly prescribed, it fails that test.

For people who have had their thyroid removed, the evidence shows that levothyroxine monotherapy often leaves them with a different problem. A 2021 study of 30 thyroidectomized patients found that after a year on levothyroxine, TSH was normal but FT3 levels were significantly lower than before surgery, while FT4 was higher [5]. They gained weight, their insulin resistance (HOMA-IR) increased, and they still reported hypothyroid symptoms [5]. This is a known limitation of giving only T4 and relying on the body to convert it to the active T3. Some people do not convert efficiently, and the standard blood test (TSH) misses this.

A 2013 crossover study found that nearly half of patients preferred desiccated thyroid extract (which contains both T4 and T3) over levothyroxine, and those on the extract lost about 3 pounds [6]. A 2025 case series reported that adding liothyronine (T3) to a reduced levothyroxine dose raised T3 levels and resolved fatigue in three patients who had persistent symptoms despite normal TSH [8]. These are small studies, but they point to a real subgroup that the one-size-fits-all T4-only approach fails.

The evidence also shows that levothyroxine absorption is easily disrupted. A 2013 crossover study found that taking it with ciprofloxacin reduced absorption by 39%, while rifampin increased it by 25% [10]. The TICO study showed that a liquid formulation can be taken with breakfast without losing effectiveness, which matters for adherence [9]. But the broader point is that the drug's bioavailability depends on gut health, timing, and other medications in ways that are rarely accounted for in routine care.

There is also a documented link between hypothyroidism and small intestinal bacterial overgrowth (SIBO). A 2007 case-control study found that 54% of hypothyroid patients had SIBO versus 5% of controls, an odds ratio of 22 [4]. Treating the SIBO with rifaximin improved GI symptoms but did not change thyroid hormone levels, suggesting the gut problem is a consequence of the hypothyroidism, not a cause [4]. A 2023 study in pregnant women with subclinical hypothyroidism found that a 21-day course of probiotics and prebiotics reduced TSH, improved cholesterol, and lowered inflammation markers [7]. This raises the question of whether addressing gut health could reduce the need for thyroid medication in some people, but the evidence is still early.

My call: for overt hypothyroidism with clear symptoms, levothyroxine is a genuine replacement therapy and can be lifesaving. For subclinical hypothyroidism, especially in older adults, the evidence shows it does not improve how people feel, and the practice of treating a lab number should stop. For anyone on levothyroxine who still feels unwell, the possibility of low T3 conversion, gut absorption issues, or SIBO should be investigated, not dismissed. The burden of proof for prescribing it to an asymptomatic person with a mildly elevated TSH has not been met.

Confidence: high for the lack of benefit in subclinical hypothyroidism in older adults; moderate for the limitations of T4-only therapy in thyroidectomized patients; low for the broader claims about gut health and alternative treatments, where the evidence is still thin.