Refractory Hypothyroidism: Chasing the Dose or Chasing the Cause
A man's levothyroxine dose has nearly tripled over three years without his TSH ever normalizing. The disagreement is whether to keep adjusting the dose and its timing, or stop and formally investigate why his gut isn't absorbing it.
Franklin O., a 59-year-old long-haul truck driver, has had his levothyroxine dose raised four times since his hypothyroidism was diagnosed three years ago — from 75 to 200 micrograms a day — and his TSH has never once come back normal, holding stubbornly between 9 and 14 μIU/mL through every increase. He takes the medication first thing in the morning, he says, usually in the cab of his truck before pulling out of the depot, with the coffee he's had every morning for twenty years. He was started on omeprazole for reflux around the same time his levothyroxine requirements first began climbing, a detail that hadn't previously been connected to his thyroid labs in his chart. He denies any diarrhea, weight loss, or abdominal pain, and his most recent complete blood count and iron studies are unremarkable.
Three plausible stories fit the same rising-dose pattern, and they point toward very different next steps. Coffee taken close to his levothyroxine dose measurably reduces its absorption — Benvenga and colleagues demonstrated this directly, and it happens independently of any gut pathology. So does the omeprazole: Centanni and colleagues showed in the New England Journal that raising gastric pH impairs the acidic environment levothyroxine tablets need to dissolve properly, pushing up the dose required to hold a given TSH. Both are fixable by separating the timing, without needing to prove anything about his intestine. But a persistently elevated TSH despite an aggressively escalated dose, in a patient with no clear timing explanation volunteered until specifically asked, is also the classic presentation of true malabsorption — celiac disease and atrophic autoimmune gastritis both travel with autoimmune thyroid disease more often than chance would predict, and neither has been ruled out. The dose-and-timing story and the malabsorption-workup story aren't mutually exclusive, but treating this as settled by the coffee-and-PPI explanation risks missing a second, independently morbid diagnosis hiding behind a plausible first one.
Deciding whether to fix the timing or investigate the gut
I'd fix the obvious interactions first. Coffee taken within an hour of levothyroxine reduces its absorption in a dose-dependent way Benvenga's group demonstrated directly, and the omeprazole compounds it — Centanni showed raised gastric pH slows tablet dissolution enough to raise the dose requirement measurably. Have him take it with water, alone, and wait sixty minutes before coffee or the PPI. That alone could explain his entire pattern without invoking anything wrong with his intestine.
I'd want celiac serology and consideration of H. pylori testing before we call this solved by timing alone. Autoimmune thyroid disease and celiac disease share genetic risk and travel together more than coincidence would explain, and atrophic gastritis — which can also impair levothyroxine absorption by the same low-acid mechanism as his PPI — wouldn't necessarily announce itself with diarrhea or weight loss the way people expect.
The timing fix is reasonable and cheap, but treating it as the explanation before checking serology risks discharging him with an unaddressed second diagnosis just because the first one was easier to find and fits the story he happened to volunteer.
There's a way to test both explanations in parallel rather than sequentially arguing about which one to chase first: a short course of directly observed, supervised weekly dosing. If his TSH drops substantially on a supervised schedule using the same weekly total dose, that confirms the problem was administration — timing, adherence, or both — not true malabsorption, and celiac serology becomes a lower-yield next step rather than a default one. If it doesn't move, that's a real, mechanistic argument for pursuing the gastroenterology workup in earnest rather than continuing to escalate his outpatient dose indefinitely.
Agreed: retime his levothyroxine away from coffee and omeprazole, and run a short supervised weekly-dosing trial with a TSH recheck before deciding on further workup.
Not agreed: whether to send celiac serology today regardless of the trial's outcome. The gastroenterologist would send it now, arguing the prevalence link with autoimmune thyroid disease justifies checking regardless of what the dosing trial shows; the endocrinologist would rather wait for the trial result to avoid an unnecessary workup if timing turns out to fully explain his pattern.