Any systemic illness suppresses total T4. This is not a hypothesis — it is a well-documented physiological response called euthyroid sick syndrome (ESS), also referred to as non-thyroidal illness syndrome (NTIS). The result: a thyroid panel drawn during an unrelated illness reads like hypothyroidism even when the thyroid is functioning normally. The dog is not hypothyroid. The dog is sick. And treating sick dogs with levothyroxine for a condition they do not have is an error that is made more often than the veterinary literature tends to acknowledge.
The Mechanism
During systemic illness, inflammatory cytokines — particularly IL-6, TNF-alpha, and IL-1 — suppress hypothalamic TRH production, pituitary TSH release, and peripheral T4-to-T3 conversion. This is an adaptive response. Reducing metabolic rate during illness conserves energy for immune response and tissue repair. Evolution built this in intentionally.
The practical consequence is that total T4, free T4, and T3 all fall during any significant systemic illness. The magnitude of suppression correlates with illness severity. A dog with mild gastroenteritis might show a T4 10-15% below its personal baseline. A dog with septic peritonitis might show T4 values that fall entirely below the reference range.
TSH in ESS is usually low-normal or normal, not elevated — which is the distinguishing feature from true primary hypothyroidism, where TSH rises in response to inadequate thyroid hormone production. But canine TSH assays have significant limitations at the low end of the range, and a normal-to-low TSH in a sick dog is easy to overlook if the T4 is the number drawing attention.
How Common Is This Problem?
A 2019 study in the Journal of Veterinary Internal Medicine evaluated 312 dogs with confirmed non-thyroidal illness and found that 34% had total T4 values below the reference range. In a companion analysis of dogs referred for hypothyroidism workup, 27% of those with low T4 were ultimately diagnosed with ESS rather than hypothyroidism after illness resolution and repeat testing.
That figure — 27% of low T4 presentations in a referral population being ESS rather than true hypothyroid — should reframe how clinicians approach the initial evaluation. A single low T4 in any dog with concurrent illness, weight loss, or recent surgery is not sufficient to initiate lifelong hormone replacement therapy.
Drugs That Make This Worse
Several commonly used veterinary medications independently suppress thyroid markers:
- Glucocorticoids: Prednisone and dexamethasone suppress total T4 by 20-40% at therapeutic doses. A dog on prednisone for immune-mediated hemolytic anemia may show T4 values that look like hypothyroidism while the thyroid axis is entirely normal.
- Sulfonamides: Trimethoprim-sulfamethoxazole and other potentiated sulfonamides inhibit thyroid peroxidase and can suppress T4 within 4-6 weeks of use. This effect is reversible on discontinuation but mimics true hypothyroidism on a panel.
- Phenobarbital: Long-term phenobarbital increases T4 metabolic clearance, leading to lower total T4 readings in epileptic dogs that may be entirely euthyroid. TSH is typically not elevated in phenobarbital-suppressed dogs, but the T4 pattern can prompt unnecessary levothyroxine initiation.
- NSAIDs: Moderate suppressive effect on T4, clinically significant mainly when combined with other confounds.
When to Test and When to Wait
The practical guideline is straightforward: do not run a thyroid panel to diagnose hypothyroidism when any of the following are present:
- Active systemic illness (fever, elevated acute phase proteins, leukocytosis)
- Current glucocorticoid therapy
- Current sulfonamide therapy within the past 8 weeks
- Recent surgery within the past 4 weeks
- Documented chronic inflammatory disease (IBD, polyarthritis)
Wait until the dog is clinically stable and off confounding medications for at least 6 weeks before running a diagnostic thyroid panel. If thyroid testing is genuinely urgent despite these confounds, include TSH, free T4 by equilibrium dialysis, and ideally a thyroglobulin autoantibody titer — the combination provides substantially more interpretive information than total T4 alone.
Free T4 by Equilibrium Dialysis
Free T4 measured by equilibrium dialysis (fT4 ED) is significantly more resistant to the suppressive effects of NTI than total T4. It is not immune — free T4 also falls in severe illness — but the suppression is less pronounced, and the assay is less affected by protein binding changes that artificially alter total T4 values.
The trade-off is cost and turnaround. Equilibrium dialysis is a reference lab procedure — it is not available on point-of-care analyzers and typically adds $40-65 to the panel cost. For a dog with a genuinely ambiguous presentation, that cost is justified. For a routine wellness thyroid screen in a healthy dog, total T4 remains appropriate as the first-line test.
Where Machine Learning Helps
Our platform flags samples for potential ESS interference based on concurrent markers in the same panel. A dog presenting with elevated CRP, low albumin, and a below-reference T4 receives a notation that the T4 result is potentially confounded by active systemic inflammation — along with a recommendation to retest after resolution rather than initiate thyroid supplementation.
This context layer is something standard reference lab reports do not provide. A T4 result delivered in isolation, without inflammatory marker context, forces the clinician to manually connect the dots. When those dots are in the same sample, connecting them automatically reduces unnecessary diagnoses.
A dog that does not have hypothyroidism should not spend the rest of its life on levothyroxine. Getting the timing and context of thyroid testing right is how that outcome is avoided.