Lithium-Induced Hypothyroidism: Fixing the Thyroid or Fixing the Regimen
A man's mood has been stable on lithium for a decade, and his thyroid has just stopped keeping up. The disagreement is whether to treat the thyroid and keep the mood stabilizer that's working, or use this as the moment to move him off a drug he's needed for years.
Desmond A., a 52-year-old sound engineer, has been stable on lithium for bipolar I disorder for eleven years, after two hospitalizations in his thirties before lithium finally brought his mood swings under control. His psychiatrist has described him as her most stable patient with his diagnosis, and he has held the same job at the same studio for nine of those eleven years. Routine labs at his annual physical this month showed a TSH of 11.4 μIU/mL with a low-normal free T4 and a mildly enlarged, non-nodular goiter on exam that wasn't there at his visit two years ago. He denies any new fatigue he's noticed himself, though his wife mentioned he's been napping most afternoons, which he'd attributed to getting older.
Lithium's effect on the thyroid is well characterized and common enough that it isn't a diagnostic surprise — McKnight and colleagues' Lancet meta-analysis of lithium's toxicity profile put hypothyroidism among its best-established long-term effects: it inhibits thyroid hormone release and can impair synthesis as well, producing both goiter and hypothyroidism in a meaningful minority of patients on long-term therapy, with risk accumulating the longer treatment continues rather than concentrating early. The straightforward fix — adding levothyroxine while continuing lithium — is also the standard one; lithium- induced hypothyroidism is not typically a reason to abandon an otherwise effective mood stabilizer, and switching psychiatric medications in a patient this stable carries its own well-documented risk of destabilizing a mood disorder that took two hospitalizations to bring under control the first time. The harder question isn't whether levothyroxine works here — it does — it's whether eleven years of accumulated lithium exposure, now producing its first visible thyroid effect, is itself a signal worth reassessing his broader treatment plan, independent of whether today's specific fix is correct.
Deciding how much of the plan to change for a well-understood side effect
I'd start levothyroxine and keep lithium unchanged. This is a well-characterized, expected side effect at his duration of exposure, not a reason to reconsider the underlying drug — thyroid replacement fully addresses it in the overwhelming majority of patients, and there's no clinical reason to complicate a stable psychiatric regimen to solve a problem that has its own straightforward fix.
I agree completely on not switching his mood stabilizer — I want to be unambiguous about that. Eleven years of stability after two hospitalizations is not something to risk on a thyroid finding that has a known, effective fix.
Where I'd add something is that "the fix works, so there's nothing else to discuss" undersells what this finding actually represents — it's the first visible sign of eleven years of cumulative lithium exposure doing something measurable to his body, and that's worth using as a prompt to review his broader monitoring plan, not just his thyroid labs specifically.
I'd start the levothyroxine today and use this visit to do exactly what's being proposed: review his renal function, since the same meta-analysis flagged reduced urinary concentrating ability and a real if modest fall in glomerular filtration as lithium's other major long-term concern, effects that also accumulate with duration rather than announcing themselves early, and confirm his last lithium level and renal panel weren't overdue. None of this changes today's thyroid decision. It's simply using a finding that prompted closer attention to make sure nothing else has been quietly accumulating alongside it.
Agreed: start levothyroxine, continue lithium unchanged, and use this visit to confirm his renal function and lithium level monitoring are current rather than overdue.
Not agreed: how much this finding should change his routine monitoring interval going forward. The psychiatrist would tighten his thyroid-function monitoring interval permanently, treating this as evidence his individual risk is higher than the general lithium population's; the endocrinologist would return him to the standard annual interval once his levothyroxine dose is stabilized, seeing no basis yet for assuming his ongoing risk trajectory differs from any other long-term lithium patient's.