Why one eye looks different from the other in thyroid eye disease

Almost every week a patient hands me a phone and asks me to look at a photograph from two years ago. Not because the picture is flattering. Because in it the two eyes match, and today they do not. One lid sits higher. One eye looks bigger, or more startled, or simply more open. The other side looks the way both sides used to look.

In thyroid eye disease that is the usual pattern rather than the odd one. The two orbits are separate compartments, and the disease does not audit itself for fairness. Working out what the difference between your two sides does and does not mean settles most of the anxiety, and occasionally it is the thing that sends me straight to a scanner.

Why does one of my eyes look bigger than the other?

Because thyroid eye disease frequently does not affect the two eyes equally, and asymmetrical disease is a common way for it to present. What you are seeing is usually not one healthy eye and one diseased eye. It is one orbit further along than the other.

Two separate things create the look, and they are worth separating in your own mind before you see anyone. The first is eyelid retraction, where the upper lid sits higher than it should and exposes a strip of white above the coloured part of the eye. Upper eyelid retraction is present in around 90 to 98 per cent of cases of thyroid eye disease, and lower lid retraction in up to 49 per cent, which makes it the commonest single reason two sides stop matching. The second is proptosis, the eye itself pushed forward out of the socket by swollen muscle and fat behind it. A retracted lid and a protruding eye look similar in a mirror and are not the same problem.

There is a third sign patients rarely notice and I always look for. Ask someone with thyroid eye disease to follow a finger slowly downwards, and the upper lid trails behind the eye instead of travelling down with it. That is eyelid lag. You cannot feel it. It shows itself in about ten seconds in a consulting room.

If only one eye looks affected, is the other one fine?

Usually not, and this is the single most useful thing I can tell you today. When orbital imaging is done in cases that appear unilateral, it confirms asymmetric changes in the extraocular muscles on both sides. The quiet eye is generally involved too. It is simply showing less.

That changes how the disease is followed. Both orbits get measured, both get examined, and neither is written off as the normal one to compare against. It also changes what you should expect over time. The side that looks untouched today may be the side that changes next, and that is not a new disease starting, it is the same one moving. I do not plan treatment for one eye as though the other were a bystander.

Is it my eyelid sitting too high, or my eye pushing forward?

They are measured differently, and the measurement is the answer. Eye position is recorded with a Hertel exophthalmometer, an instrument that rests on both outer orbital rims and reads, in millimetres, how far the front of the cornea sits ahead of the line drawn between the two cheekbones. What matters in clinic is your own pair of numbers, the gap between your two sides, and how both behave over months.

Where the two sides are far apart, or where the picture does not look like textbook thyroid disease, computed tomography based exophthalmometry is described as the gold standard for measuring proptosis, because it is accurate, reproducible, and not thrown off by swelling of the tissue around the eye. Eyelid position is judged separately, by where the lid margin sits against the cornea and how much white shows above it.

Telling those two apart matters for more than tidiness. A lid that will not come down and an eye that sits forward both leave corneal surface uncovered, and the cornea is the part of this disease that can quietly cost you sight. I have written separately about eyes that will not fully close at night, which is where asymmetry stops being something you notice in photographs and becomes a surface problem.

When does a one-sided eye need a scan?

When the presentation is atypical or asymmetrical, imaging earns its place. Magnetic resonance imaging of the orbits is useful in exactly that situation, to exclude the other conditions on the list rather than to confirm what is already obvious. That is not the same as saying every asymmetric eye needs a scan, and plenty of patients are safely followed without one.

The reasoning behind it goes like this. Exophthalmos means protrusion of one eye or of both. In adults, thyroid associated eye disease is the most common cause of proptosis overall, and yet it accounts for fewer than a third of cases where only one eye protrudes. So most one-sided proptosis in an adult has another cause: orbital cellulitis or another inflammatory condition such as sarcoidosis, a benign or malignant orbital tumour with lymphoma among them, a carotid cavernous fistula or orbital varices, or bleeding into the socket after a fracture of the orbital bones.

The features that make me stop assuming thyroid are worth committing to memory: rapid onset, pain and loss of vision. A rapid onset points towards inflammatory disease, a malignant tumour or a carotid cavernous fistula, where a gradual one usually implies something more benign. Pain most commonly points to infection. Loss of vision raises the possibility of swelling pressing on the optic nerve, which is rare and is exactly why nobody waits on it. Separating orbital inflammation from lymphoma before anyone picks up a biopsy forceps is genuinely difficult, and it is a question my colleagues and I have studied and presented, listed under my research. What else lives in the socket, and how each thing announces itself, is set out in my piece on orbital conditions.

Does the worse-looking side mean my disease is severe?

No. How different the two sides look is not how activity or severity is graded, and I would not want anyone reading their own face that way. Activity is scored formally. Under the amended Clinical Activity Score, disease is called active when the score is above 3 out of 7 at a first visit, or above 4 out of 10 at later visits, and that score measures inflammation, not symmetry.

Two things do change the course. Smoking is strongly associated with increased severity of thyroid eye disease and is the most significant modifiable risk factor in it, which makes stopping the one instruction here that costs nothing. Radioiodine treatment for Graves disease also raises the risk of orbitopathy, and that is a conversation to have with your endocrinologist before the treatment rather than after it. Symptoms usually last one to two years and often settle on their own, though settling is not the same as returning to the face you had before. Controlling activity during that window, on as little steroid as the disease will allow, is the part I spend most of my clinic time on, and it is the subject of a 14-year retrospective cohort I presented with colleagues from a tertiary oculoplastic service.

My thyroid blood tests are normal. Can this still be thyroid eye disease?

Yes, and this catches people out constantly. At diagnosis, 90 per cent of patients with thyroid eye disease have an overactive thyroid, 6 per cent have normal thyroid function, 3 per cent have Hashimoto thyroiditis and 1 per cent have an underactive thyroid. A normal thyroid blood test does not exclude the disease, and an abnormal one does not by itself prove that what is happening to your eye is thyroid related.

One result does track with how the eye behaves. Thyrotropin receptor antibody levels correlate with severity, which is why that antibody is often more interesting to me than the hormone numbers, and why I ask for it even in patients whose thyroid function has been called normal for years. The wider picture of what this disease does and what treatment involves sits in my longer piece on the signs and treatment of thyroid eye disease.

What is actually done about the difference between my two eyes?

The order is the answer: activity first, appearance later. While the disease is active, the work is protecting the eye surface, controlling inflammation and watching the optic nerve, because rarely the swelling around the eye presses on that nerve and causes vision loss. Surgery to reposition a lid or to decompress a socket is planned once the disease has gone quiet, and it is planned across both sides even when only one is obviously off, since operating on the louder eye alone tends to make the difference more visible rather than less.

Between those two phases there is a stretch of months where the honest answer is that we measure, we protect and we wait. I know how unsatisfying that is to read at midnight. If you have arrived here because your face has changed and you are trying to work out what is ageing and what is not, my piece on what makes eyes look tired comes at the same question from the other end. For the anatomy of the socket itself, start with the orbit and thyroid eye disease section.

Asymmetry is not a verdict. It is information: about which orbit is further along, about whether the story fits thyroid disease at all, and about how closely the quieter side needs watching. Bring the old photographs. They are better evidence than anybody's memory, mine included.