GLP-1 Receptor Agonists and Thyroid Function: MTC Risk Context, Levothyroxine Pharmacokinetics, and Anti-inflammatory Effects in Hashimoto's

[Image: Immunohistochemical comparison of GLP-1R expression in rat vs. human thyroid C-cells]

The mechanistic basis for the rodent MTC signal is well-characterized: rat and mouse thyroid C-cells express GLP-1R at high density. GLP-1RA-mediated GLP-1R activation on C-cells stimulates calcitonin secretion and, at sustained high-dose exposure, drives C-cell proliferation via cAMP/PKA and downstream ERK1/2 signaling. At lifetime carcinogenic dose exposures in rodents, this proliferative stimulus resulted in C-cell hyperplasia and frank MTC in male and female rats and mice across multiple GLP-1RA compounds — a class effect attributable to shared GLP-1R-mediated mechanism.

The translational limitation is the fundamental difference in human thyroid C-cell GLP-1R expression. Multiple immunohistochemical and mRNA expression studies of human thyroid tissue have consistently shown either absent or very low GLP-1R expression in parafollicular C-cells. A 2011 study by Knudsen et al. (Endocrinology) directly examined human thyroid tissue versus rat thyroid tissue using validated GLP-1R antibodies and found substantial GLP-1R staining in rat C-cells with minimal to absent signal in human C-cells. This finding has been replicated in multiple independent tissue expression datasets. The molecular substrate for rodent GLP-1RA-driven C-cell proliferation simply does not exist in humans at the same receptor density.

Post-marketing pharmacovigilance supports this mechanistic prediction. A 2023 retrospective cohort study using Danish health registry data (n=145,316 GLP-1RA users vs. matched controls) found no significant increase in MTC or thyroid cancer incidence with GLP-1RA exposure over a median follow-up of 5.5 years. A FAERS (FDA Adverse Event Reporting System) analysis showed no disproportionality signal for MTC versus other diabetes drug classes. Long-term cardiovascular outcomes trials (LEADER, SUSTAIN-6, REWIND) with 2-5 year follow-up in large populations similarly found no MTC excess.

The clinical contraindication remains valid and important: patients with personal history of MTC, familial MTC, or MEN2A/2B syndrome should not use GLP-1RAs. In these patients, any marginal C-cell proliferative stimulus — even below the threshold detectable in general population studies — is unacceptable given the existing genetic risk. The contraindication is not evidence that the drug causes MTC in average-risk individuals; it is a precautionary exclusion for a population where even low additional risk is unjustifiable.

Levothyroxine (T4) is a narrow therapeutic index drug with absorption characteristics that are highly sensitive to gastric emptying rate and luminal conditions at the time of ingestion. Standard dosing protocol — administration in a fasting state 30-60 minutes before food — is designed to maximize duodenal absorption before food, calcium, iron, coffee, or fiber interfere with the T4 molecule. GLP-1RAs introduce a pharmacokinetic perturbation to this carefully managed absorption window through their dose-dependent delay of gastric emptying.

GLP-1RA-mediated gastric motility inhibition operates via GLP-1R activation on enteric neurons, reducing acetylcholine release at the neuromuscular junction of the gastric antrum and pylorus. This delays the gastric emptying half-time (T50) — the time for 50% gastric contents to empty into the duodenum — by approximately 70-120 minutes at therapeutic semaglutide doses in pharmacoscintigraphy studies. For levothyroxine taken in the morning with water on an empty stomach, the drug should theoretically transit to the duodenum rapidly under normal conditions. However, if a GLP-1 injection the previous day or week has altered basal gastric tone, or if the levothyroxine is taken within hours of the GLP-1 injection, delayed transit could reduce peak duodenal T4 concentration and lower total absorption (AUC).

The clinical consequence is a reduction in effective levothyroxine bioavailability, manifesting as a rise in TSH (indicating hypothyroid underdosing) without any change in dose. This effect is analogous to — though mechanistically distinct from — the levothyroxine-calcium carbonate and levothyroxine-iron interaction, both of which are well-documented causes of iatrogenic hypothyroidism through absorption interference. The practical management protocol involves temporal separation: levothyroxine in the morning on awakening with water only, and GLP-1 injection on a different day (for weekly dosing, typically the evening of the preferred injection day) to minimize overlap in gastric motility effects. Tirosint liquid gel capsules may offer slightly better absorption robustness due to the gel formulation bypassing some luminal variable absorption, but this is not established in GLP-1RA co-administration studies.

TSH monitoring at 6-8 week intervals during GLP-1RA initiation and during the weight loss phase is the standard of care in patients on levothyroxine. Metabolic rate changes accompanying weight loss also shift levothyroxine requirements downward in some patients over time, creating a second mechanism for dose adjustment needs independent of absorption changes.

Hashimoto's thyroiditis is characterized by CD4+ T helper cell infiltration of the thyroid gland (predominantly Th1 phenotype), B-cell production of anti-thyroid antibodies (TPO-Ab, Tg-Ab), and a sustained inflammatory cytokine environment (IL-1β, TNF-α, IFN-γ, IL-6) that drives both local thyroid tissue destruction and systemic immune dysregulation. The nuclear factor kappa-B (NF-κB) transcription factor complex is a central regulator of this cytokine production, activating the promoter regions of IL-1β, TNF-α, IL-6, and multiple chemokines that perpetuate the autoimmune infiltrate.

GLP-1R activation has been consistently shown to inhibit NF-κB signaling through cAMP/PKA-mediated phosphorylation of IκBα (inhibitor of κB alpha), preventing IκBα degradation and consequently blocking the translocation of NF-κB p65/p50 heterodimers to the nucleus. This anti-inflammatory mechanism is documented in multiple tissue types including pancreatic beta cells, macrophages, vascular endothelium, and neuronal cells — and is believed to represent a class-wide GLP-1RA effect rather than a molecule-specific property. The downstream consequence in Hashimoto's tissue would be reduced cytokine-driven thyroid follicular cell apoptosis and reduced autoimmune infiltrate sustaining signals.

Direct clinical evidence for GLP-1RA-mediated TPO antibody reduction is early-stage. A 2025 Italian study (n=88) examining Thymosin Alpha-1 — a peptide with analogous anti-inflammatory NF-κB-inhibitory properties — demonstrated a statistically significant reduction in TPO-Ab titers over 6 months of treatment in Hashimoto's patients. This provides indirect mechanistic support for GLP-1RA-mediated effects, but is not GLP-1RA data. Published GLP-1RA case series in Hashimoto's patients are small (n=5-20) and report variable TPO-Ab reduction, insufficient to establish a treatment effect. A prospective investigation specifically in Hashimoto's patients is needed.

The indirect pathway — GLP-1RA-mediated weight loss reducing systemic inflammation — may be equally important. Obesity is associated with elevated circulating IL-6, CRP, and adipokine dysregulation (leptin excess, adiponectin deficiency) that amplifies autoimmune inflammation in susceptible individuals. Lean mass-to-fat mass ratio improvement with GLP-1RA treatment shifts this systemic inflammatory tone, potentially reducing the substrate that sustains Hashimoto's autoimmune activity. This mechanism predicts that TPO-Ab reduction, if observed, would be greatest in patients with higher pre-treatment BMI and adipose-driven inflammation.

The T4-to-T3 conversion pathway is mediated by deiodinase enzymes — type I deiodinase (DIO1) in liver and kidney, type II deiodinase (DIO2) in muscle, brain, and thyroid, and type III deiodinase (DIO3) inactivating both T4 and T3. Adipose tissue expresses DIO2 at low but functionally significant levels, contributing to peripheral T3 production. Body composition changes accompanying GLP-1RA treatment — specifically the reduction in adipose mass and the potential increase in lean mass-to-fat ratio — can modulate the relative contribution of different tissue compartments to total T4-to-T3 conversion, shifting the equilibrium of thyroid hormone activation.

In clinical practice, this manifests as TSH-normal patients reporting hypothyroid symptoms (fatigue, cognitive slowing, cold intolerance) during the active weight loss phase — symptoms consistent with transient shifts in free T3 availability rather than inadequate T4 dosing. The conversion shift is typically self-limiting as body composition stabilizes, but monitoring free T3 (in addition to TSH and free T4) during the active weight loss phase can identify patients with clinically significant conversion impairment requiring temporary dose adjustment. This monitoring is not yet standard of care but is mechanistically justified for patients on levothyroxine monotherapy who may not generate adequate free T3 from T4 through conversion alone.

Autonomic changes accompanying GLP-1RA treatment also affect thyroid function testing interpretation. GLP-1R activation has sympathomimetic-attenuating effects, reducing norepinephrine-driven thyroid stimulation. Combined with the metabolic rate reduction inherent to significant weight loss, TSH set-point may shift slightly upward even in euthyroid (not hypothyroid) individuals — not because thyroid function is impaired but because the metabolic demand has changed. This makes the interpretation of borderline TSH elevations during GLP-1RA weight loss more nuanced: a TSH of 3.8 mIU/L during rapid weight loss may represent appropriate physiological re-calibration rather than levothyroxine under-dosing. Clinical judgment integrating symptoms, free T4, and free T3 is required.

For patients with Hashimoto's on combination T4/T3 therapy or NDT (natural desiccated thyroid) preparations, the dosing adjustment considerations are the same but more complex given the fixed T4:T3 ratio in NDT products versus the adjustable flexibility of levothyroxine + liothyronine. Patients on these regimens initiating GLP-1RA therapy should have more frequent monitoring (every 4-6 weeks initially) and a lower threshold for consultation with a thyroid-specialist endocrinologist.