How Is Thyroid Eye Disease (TED) Diagnosed?

Thyroid eye disease is diagnosed through a combination of clinical examination, blood tests for thyroid-related antibodies, and often orbital imaging. There is no single test that confirms or rules it out. Instead, doctors piece together findings from the eyes, the thyroid gland, and sometimes a CT or MRI scan to arrive at a diagnosis and determine how active and severe the disease is. That layered approach matters because TED can look different from person to person and can even show up before any thyroid blood work comes back abnormal.

What Doctors Look for During the Eye Exam

The physical examination is the backbone of a TED diagnosis. An ophthalmologist or endocrinologist will check for several hallmark signs. Eyelid retraction, where the upper eyelid sits higher than normal and exposes more of the white of the eye, is the most common finding. Proptosis, the medical term for eyes that bulge forward, is another classic sign. Doctors measure proptosis with a small ruler-like instrument called an exophthalmometer, which gives a millimeter reading of how far each eye projects from the bony rim of the eye socket.

Beyond those two big signs, the examiner looks at eyelid swelling, redness or puffiness of the tissue around the eyes, redness over the eye muscles (visible on the white of the eye), restriction of eye movement, and whether the eyelid lags behind when you look downward. Each of these signs can be graded, and together they paint a picture of how much the disease has affected the orbit.

Eye movement is tested carefully because the muscles that rotate the eyes are a primary target of the disease. When those muscles become inflamed and swollen, they lose their flexibility. You might notice double vision when looking in certain directions, or the doctor might find that one eye cannot rotate as far as the other. Motility testing, sometimes aided by a tool called a Hess chart, maps exactly which muscles are restricted and how severely.

The Clinical Activity Score

Once TED is suspected, doctors need to determine whether the disease is currently active (inflamed and still changing) or inactive (burned out and stable). This distinction drives treatment decisions because immunosuppressive therapies work best during the active phase. The most widely used tool for this is the Clinical Activity Score, a checklist of seven inflammatory signs scored as present or absent. The signs include spontaneous eye pain, pain on eye movement, redness of the eyelids, redness of the conjunctiva (the white part of the eye), swelling of the eyelids, swelling of the caruncle or plica (small fleshy structures near the inner corner of the eye), and chemosis (swelling of the conjunctiva itself).

A score of three or more out of seven is generally considered active disease. Research has shown that patients with a higher pretreatment score are more likely to respond to anti-inflammatory treatment. In one study, a cutoff of four or higher predicted a good treatment response with about 86 percent specificity and 80 percent positive predictive value, though its sensitivity was more modest at 55 percent.¹ That means a high score is a fairly reliable green light for treatment, but a lower score does not necessarily mean the disease is quiet.

One acknowledged limitation of the CAS is its subjectivity. Two doctors examining the same patient may score some signs differently, particularly the softer tissue signs like mild eyelid redness or subtle chemosis. Efforts to standardize the process include photographic scoring, where periocular photographs are assessed remotely, though studies have found that photo-based scores perform slightly less well than in-person clinical scoring unless demographic information is also factored in.² The CAS remains the standard, but clinicians treat it as one input among many rather than a definitive verdict on its own.³

Blood Tests and Antibody Levels

Most people diagnosed with TED have Graves’ disease, an autoimmune condition where the immune system produces antibodies that stimulate the thyroid. Blood tests for thyroid function (TSH, free T4, and sometimes free T3) help establish whether a thyroid disorder is present. But for diagnosing TED specifically, doctors look at thyroid-stimulating immunoglobulins (TSI) and thyrotropin receptor antibodies (TRAb).

TSI are antibodies that mimic thyroid-stimulating hormone and activate the TSH receptor. In TED, these same antibodies also target tissues behind the eye, which is why they are diagnostically useful. One study found that a newer TSI assay had 93 percent sensitivity and 100 percent specificity for TED diagnosis compared with healthy controls.⁴ TSI levels also correlate with disease activity and severity: higher levels tend to track with more active, more severe eye disease. Research has identified a specific TSI cutoff level that predicts active TED with about 77 percent sensitivity and 81 percent specificity.⁵

That said, the picture is not perfectly clean. Some research has found that TSI and TRAb levels do not always differ significantly between Graves’ patients who develop eye disease and those who do not, particularly in smaller cohorts followed over time.⁶ So while a strongly positive TSI result in someone with suggestive eye findings clinches the diagnosis, antibody levels alone cannot predict with certainty who will or will not develop TED.

When Normal Thyroid Levels Complicate the Diagnosis

A common misconception is that you cannot have TED unless your thyroid is clearly overactive. In reality, a small but meaningful fraction of TED patients are euthyroid, meaning their thyroid hormone levels are entirely normal. This makes diagnosis trickier because the usual red flag of abnormal thyroid function is absent. Euthyroid Graves’ disease shares many features with the classic form but can go unrecognized longer because doctors may not think to look for it.⁷

In these cases, the eye findings themselves are what prompt the workup. Antibody testing becomes especially important: a positive TRAb or TSI result in a patient with normal thyroid levels and suggestive eye signs points strongly toward TED. Orbital imaging helps too, because the pattern of muscle enlargement seen on CT or MRI is fairly distinctive. A systematic review emphasized that clinicians should remain aware that TED can show up with normal thyroid function and recommended a full workup including orbital images, thyroid hormone levels, and antibody testing in any suspicious case.⁸ Eye disease can even be the very first sign of Graves’ disease, appearing months or years before thyroid function goes haywire.

The Role of Imaging

CT and MRI scans of the orbits are not needed in every case of TED, but they become important in several situations: when the diagnosis is uncertain, when doctors suspect the optic nerve may be compressed, or when surgical planning requires detailed anatomy. Both imaging modalities show the hallmark finding of TED in the orbit, which is enlargement of the extraocular muscles. A key distinguishing feature is that in TED, the muscle bellies swell while the tendons are spared. This pattern helps separate TED from other conditions like orbital pseudotumor, where the inflammation tends to involve the tendon as well.⁹

MRI has an additional advantage: it can detect active inflammation in the muscles using specialized sequences. T2-weighted images and T2-relaxation mapping pick up edema (water content) in inflamed muscles, which appears bright on the scan. Research has shown that T2-relaxation times clearly distinguish between normal and inflamed eye muscles, and these measurements actually drop as patients respond to treatment, sometimes even before the clinical signs improve.¹⁰ Newer MRI techniques also measure the fat fraction within the muscles, which increases in TED and may help predict which patients will respond to steroid therapy.¹¹

Color Doppler ultrasound offers yet another window. It measures blood flow through the orbital vessels and has shown that patients with active TED have higher blood flow velocities in the superior ophthalmic vein and changes in the ophthalmic artery compared with both inactive patients and healthy controls.¹² Ultrasound is cheaper and more accessible than MRI, though it provides less anatomical detail.

Detecting Optic Nerve Involvement

The most serious complication of TED is dysthyroid optic neuropathy (DON), where swollen muscles compress the optic nerve at the back of the eye socket. This can threaten vision permanently if not caught early. Diagnosing DON involves a specific battery of tests layered on top of the usual TED workup.

Doctors check visual acuity, color vision, the pupil’s response to light (looking for a relative afferent pupillary defect, where one pupil reacts more sluggishly), and the appearance of the optic disc for swelling. Visual field testing maps out any blind spots. In one series, visual field defects were found in roughly 81 percent of eyes diagnosed with DON, and reduced visual acuity was present in about three-quarters of affected eyes.¹³

A particularly useful early warning sign is color vision loss, specifically a difficulty discriminating blue-yellow (tritan) colors. One study found that tritan deficiency, measured with a standardized color test, had a sensitivity of about 97 percent for detecting subclinical optic neuropathy, far outperforming red-green color testing.¹⁴ This matters because tritan deficiency often appears before visual acuity drops or before visual field defects become obvious, giving doctors a head start on treatment.

Imaging plays a supporting role in DON diagnosis. CT or MRI may show “apical crowding,” where the enlarged muscles pack the narrow back portion of the eye socket so tightly that the optic nerve is squeezed. Visual evoked potentials, an electrical test that measures how quickly signals travel along the optic nerve to the brain, can also pick up subtle nerve damage: reduced amplitude on this test was found in nearly 97 percent of eyes with DON in one series.¹³ A systematic review confirmed that no single test is sufficient to diagnose DON on its own; instead, a combination of clinical features and ancillary tests provides the strongest diagnostic confidence.¹⁵

Ruling Out Conditions That Mimic TED

Several other conditions can cause bulging eyes, swollen eye muscles, or orbital inflammation, and part of diagnosing TED is ruling these out. Orbital pseudotumor (idiopathic orbital inflammation) is one of the closest mimics. It can produce proptosis, pain, and restricted eye movement that look quite similar to TED. The distinguishing clues often come from a combination of antibody tests and imaging. In TED, TRAb is typically positive and the muscle bellies enlarge with tendon sparing. In pseudotumor, TRAb is negative and inflammation may involve the tendon, the lacrimal gland, or orbital fat in patterns not typical of TED. Case reports have documented patients initially treated for euthyroid Graves’ disease whose MRI findings and biopsy ultimately revealed orbital pseudotumor instead.

Other conditions in the differential include orbital lymphoma, IgG4-related disease, sarcoidosis, and orbital cellulitis. When the clinical picture is ambiguous, especially if the patient is euthyroid and antibody-negative, a biopsy of orbital tissue may be considered. Histopathology in TED tends to show a mix of inflammatory cells and glycosaminoglycan deposition in the connective tissue around the muscles, though findings can be quite variable. One study examining tissue from TED patients found that many biopsies showed only scattered inflammatory cells and noted that the inflammation may be unevenly distributed throughout the muscle, so a single biopsy sample may not always be representative.¹⁶ Biopsy is therefore a last resort rather than a routine diagnostic step.

Screening Questionnaires and Early Detection

Because TED can develop gradually and because many patients with Graves’ disease are managed by endocrinologists rather than ophthalmologists, there has been a push to create simple screening tools that catch the eye disease early. One validated symptom questionnaire, designed for patients newly diagnosed with Graves’ hyperthyroidism, asks about common TED symptoms such as grittiness, tearing, light sensitivity, double vision, and visual changes. In its validation study, it achieved a sensitivity of 76 percent and specificity of 82 percent when compared with a full orbital examination.¹⁷

A more recent telemedicine-based screening tool combined a self-reported symptom questionnaire with photographic analysis of the face. The questionnaire alone reached an area under the curve of 0.93 for detecting TED, and adding the photos nudged it slightly higher to 0.94, with sensitivity above 89 percent and specificity above 95 percent.¹⁸ Tools like these could help endocrinologists flag patients who need referral to an eye specialist without requiring a full ophthalmological setup in their clinic. Clinical guidelines already recommend that endocrinologists screen Graves’ disease patients for eye involvement using a brief focused assessment.¹⁹

TED in Children

TED is far less common in children than in adults, but it does occur and presents its own diagnostic challenges. A meta-analysis of pediatric cases found that the average age at presentation was about 12 years, with girls affected roughly twice as often as boys. The most common findings were similar to adults: exophthalmos was present in nearly all cases, eyelid retraction in about 73 percent, and dry eye symptoms in about 66 percent.²⁰ The diagnostic approach is the same in principle, but children may have trouble articulating symptoms like double vision or eye discomfort, which can delay recognition. Pediatricians who manage childhood Graves’ disease should be alert to subtle eye changes, particularly if a child complains of eye irritation or if parents notice a change in the appearance of the eyes.

Emerging Tools and Tear Fluid Biomarkers

Researchers are exploring whether tear fluid could one day supplement or simplify TED diagnosis. Tears are in direct contact with the eye surface and the lacrimal glands, which can themselves be affected by the disease, making them a potentially rich source of diagnostic molecules.²¹ A recent study found that two inflammatory markers in tear fluid, CD40 ligand and a chemokine called CCL2, were significantly higher in Graves’ disease patients with TED compared to those without eye involvement. When these tear markers were combined with a standard ocular surface discomfort score, they could distinguish TED from non-TED Graves’ patients with reasonable accuracy.²² This is still early-stage research, but the appeal is obvious: collecting a tear sample is painless and could potentially be done in any clinic.

Artificial intelligence is another frontier. Several research groups have trained deep learning models to detect TED from ordinary facial photographs. In one comparative study, two AI models screened for TED using front-facing photos and achieved area-under-the-curve values of 0.82 to 0.91 under real clinical conditions, with sensitivity reaching 88 to 92 percent.²³ Other AI models have been developed to automatically score the CAS from photographs, potentially reducing the inter-observer variability that clinicians have long acknowledged as a weakness of the scoring system. A scoping review of these AI tools concluded that they show high diagnostic accuracy across detection of presence, activity, and severity of TED.²⁴ None of these tools have replaced the clinical exam, but they could eventually serve as triage or remote-monitoring aids.

The Molecular Puzzle Behind the Diagnosis

Understanding why antibody testing works for TED requires a brief look at what is happening at the molecular level. The TSH receptor, the same receptor that antibodies target on the thyroid gland, is also present on cells called orbital fibroblasts that live in the fat and connective tissue behind the eye. When TSI antibodies bind to these receptors, they trigger inflammation, swelling, and the buildup of water-attracting molecules in the orbit. This is the core reason the eyes are affected in an autoimmune thyroid disease.

Recent research has added a wrinkle: the TSH receptor does not act alone. It physically interacts with another receptor, the IGF-1 receptor, on the surface of orbital fibroblasts. Studies using molecular modeling and laboratory experiments have demonstrated that these two receptors can form a stable complex through direct contact between their outer portions, even without the signaling intermediary that was previously thought necessary.²⁵ Blocking the IGF-1 receptor can shut down activation driven by both IGF-1 and TSH receptor-targeting antibodies.²⁶ This receptor partnership is the scientific basis for teprotumumab, a drug that blocks the IGF-1 receptor and has become a major treatment for TED. Understanding this biology also hints at why TSI levels correlate with disease activity: higher antibody levels mean more activation of this dual-receptor system.

Quality of Life Assessments as a Diagnostic Companion

While not diagnostic in the traditional sense, quality-of-life questionnaires are increasingly used alongside clinical measurements to capture how TED is actually affecting patients. The most widely used is the GO-QOL, a questionnaire with two subscales covering visual functioning and appearance-related concerns. Studies across multiple populations have confirmed that GO-QOL scores correlate with disease severity and activity: patients with moderate-to-severe or sight-threatening TED consistently report worse visual functioning and appearance scores than those with mild disease.²⁷ Patients with active disease scored particularly poorly on both subscales compared with those whose disease had stabilized.²⁸

These questionnaires serve a practical purpose in diagnosis and treatment monitoring. They quantify the patient’s experience in a way that clinical measurements like proptosis readings or CAS scores may miss. Someone with relatively modest proptosis might still be profoundly affected by dry eyes, light sensitivity, or the psychological impact of a changed appearance. GO-QOL scores have also been validated as convertible to health utility values, meaning they can be used in formal assessments of whether a treatment is producing meaningful improvement from the patient’s perspective.²⁹