Deprescribing levothyroxine is appropriate and safe in selected patients
It is chiefly used in maintaining a euthyroid status in those started on the drug for nonspecific symptoms or non–evidence-based indications, those with mild subclinical hypothyroidism, those with a suppressed TSH while on treatment, and older adults on low doses.
The core approach is gradual tapering (typically by 25 µg every 6–8 weeks) with symptom review and thyroid function testing every 6–8 weeks.
Levothyroxine is among the most prescribed medications in the US, and a large proportion of prescriptions are for mild subclinical hypothyroidism or non–evidence-based indications like euthyroid patients with weight gain or depression.
High-certainty evidence shows no benefit of thyroid hormone on quality of life or hypothyroid symptoms in subclinical hypothyroidism, yet once started it is rarely stopped.
Deprescribing aims to reduce unnecessary treatment, patient burden, potential harms-iatrogenic thyrotoxicosis, atrial fibrillation, bone loss, and cost.
Unclear or non–evidence-based original indication — patients on levothyroxine without a documented diagnosis of hypothyroidism. In a prospective cohort of 802 such patients who discontinued, only 23% became hypothyroid over a median 18 months; if hypothyroidism did not appear within the first 4 months, the probability of later progression was ~3%.
Prior subclinical (vs. overt) hypothyroidism — in a meta-analysis of 17 studies , ~37% remained euthyroid after discontinuation overall, but success was substantially higher for those originally diagnosed with subclinical hypothyroidism (35.6%) than overt hypothyroidism (11.8%).
Low baseline dose and older age — lower daily doses strongly predict successful discontinuation.
Suppressed serum TSH on treatment/overtreatment— a clear trigger to reduce or stop.
The dose-dependence trial of adults aged ≥60 years, where those on ≤50 µg/d had roughly a 64% success rate versus near or below 10% at doses above 100 µg/d.
Tapering gradually rather than abrupt cessation — commonly by 25 µg every 6–8 weeks, monitoring symptoms and rechecking TSH/free T4 at each step.
Confirm biochemical status before stopping/reducing, and after discontinuation continue periodic monitoring, since a subset will require reinitiation.
Adults with subclinical hypothyroidism on ≤75 µg/d randomized to continue vs. placebo found deprescribing feasible with no significant difference in quality of life, tiredness, weight, BMI, or lipids and no serious adverse events.
Patients with genuine overt hypothyroidism (e.g., prior thyroidectomy, radioiodine ablation, high-titer autoimmune disease with clearly elevated TSH) are poor candidates given the low likelihood of sustained euthyroidism.
Caveat on acute illness:
Not stopping or down-titrating levothyroxine on the basis of thyroid tests drawn during acute or critical illness, when non-thyroidal illness syndrome and drug effects (e.g., dopamine, glucocorticoids) distort TSH, T3, and T4.
Established replacement should be continued during acute hospitalization, with reassessment once the patient recovers.