Thyroid eye disease and the eyes that will not close properly: exposure and the cornea at risk

Most people with thyroid eye disease come to me because of how their eyes look. Staring. Bulging. Puffy by morning and never quite settling. The thing I am watching for at the slit lamp is something they have usually not mentioned and often cannot yet feel: how much of the cornea is still covered when the lids come together.

The cornea is the clear window at the front of the eye. It has no blood supply of its own, so it depends on the tear film, and the tear film is spread by a full blink. Take away the full blink and you have taken away the cornea's maintenance. That is why an eye that will not close sits in a different category from an eye that simply protrudes.

Why will my eyes not close all the way?

Because two separate changes are happening at once and they add up. Thyroid eye disease pulls the upper lid backwards, and it pushes the eyeball forwards. The lid now has further to travel over a target that has moved towards it, and sometimes it does not arrive.

The retraction is not the lid stretching. Muller's muscle, a small smooth muscle sitting behind the main lifting muscle of the upper lid, is stimulated, and the lid rests higher than it should. It characteristically pulls up most at the outer end, which is why the lid margin takes on a shallow S shape and why the sign has a name of its own: lateral flare. Behind that, the tissues inside the orbit, the bony socket that holds the eye, swell and take up room the socket does not have, so the eye is displaced forwards. That is proptosis. If you want the wider picture of how the disease starts and what else it does to the eye, I have set it out in the signs and treatment of thyroid eye disease.

Incomplete closure has its own word. Lagophthalmos means the eyelids do not close fully or do not close normally. It is not a diagnosis on its own. It is a measurement, and once it is there, the surface of your eye is exposed to the room.

What actually happens to a cornea that stays uncovered?

It dries in a sequence, and the sequence is predictable. First the tear film becomes unstable, because it is evaporating faster than a blink can replace it. Then the surface layer of the cornea starts to break down in fine dots, which is what I am looking for when I put a drop of orange dye in your eye and switch to blue light.

Those dots have a name, punctate epithelial erosions, and they are the earliest visible evidence that exposure is doing damage rather than merely causing symptoms. If it continues, the dots merge into an epithelial defect, a raw patch where the surface layer is missing altogether. A raw patch can become infected, infection in a cornea is called microbial keratitis, and keratitis can ulcerate, scar and, at the far end, perforate. Scarring in the visual axis is scarring you read on the vision chart. That is why I take a millimetre of lagophthalmos seriously in someone who is otherwise doing well.

Almost nobody travels the whole of that sequence. Most people stop at the first or second step and stay there for months. But nobody reaches the far end without passing the early steps, and the early steps are the ones findable in a clinic room in two minutes.

Why are my eyes worst first thing in the morning?

Because the exposure is happening while you are asleep and you have no idea it is happening. Lagophthalmos that occurs during sleep is called nocturnal lagophthalmos, and the person who has it is usually unaware of it. It announces itself as symptoms on waking, then quietens down through the day.

So the history I take is not really about the daytime. I ask what the first ten minutes of the morning are like. Is there a sandy feeling before you have even sat up? Does the vision take a dozen blinks to clear? Is one eye consistently worse? And if there is someone else in the house, I ask them, because a partner who has noticed a strip of white showing under your lid at night has given me better information than any symptom score.

The other thing I check is Bell's phenomenon, the reflex that rolls the eye upwards behind a closing lid. In most people it hides the cornea under the upper lid even when closure is imperfect, which is a considerable piece of natural protection. Poor Bell's phenomenon takes that protection away, so the same measured gap on two different patients is not the same amount of risk. Thyroid eye disease can also stiffen the muscles that move the eye, which is a separate problem with a separate set of consequences that I have written about under double vision that is not a squint.

What do lubricants, ointment and taping actually do?

Each does one specific job, and none of them treats the disease. They protect a surface being left uncovered, and they buy time while the underlying condition is managed. That is worth saying plainly, because patients are often handed a bottle and told nothing else.

• Artificial tears replace the film that is evaporating, and frequency is the point. A drop four times a day does very little for a surface that is drying out all day.
• Ointment is thicker and stays put, so its place is at night, or in the day if the exposure is severe. It blurs vision while it is on, which is exactly why it suits bedtime.
• Taping the eyelids at night offers additional surface protection without an operation. The National Eye Institute suggests it specifically for lids that do not close all the way.
• Moisture chamber glasses hold humid air against the eye and help keep the tear film stable.

What none of these does is close your eye. They are a dressing over an anatomical problem, and the honest framing in clinic is that they protect the surface, which is not the same as keeping it intact. Exposure after any lid procedure follows the same physics, which is why I use much of this same conversation with patients thinking about dry eye after eyelid surgery.

When does this stop being something I manage at home?

There is a real line, and the European guideline draws it in a useful place. Corneal exposure that still responds to lubricants sits in the mild category. Corneal breakdown, alongside optic nerve involvement, is what defines sight-threatening thyroid eye disease and is treated as an emergency.

Between those two points is the territory I want people to recognise early. The guideline is direct: severe corneal exposure should be treated urgently, medically or with progressively more invasive surgery, so that it does not progress to breakdown, and breakdown itself should be addressed surgically straight away. In practice that means new constant pain rather than grittiness, a white or grey patch on the clear part of the eye, light sensitivity that has changed in character, discharge, or vision that drops and does not come back after blinking. None of that is a wait-and-see list.

What surgery is there for a lid that will not close?

Several, chosen by what is causing the gap and how long it is expected to last. This is where an oculoplastic assessment earns its place, because the lid, the eye position and the surface are separate problems and they are not corrected in the same way.

Where recovery is expected within weeks, a temporary tarsorrhaphy, stitching part of the lids together to narrow the opening, protects the cornea until the situation changes. Where the upper lid is too high, lengthening it by recessing the levator and Muller's muscles lowers it back over the eye. Where the lid is heavy and slow rather than retracted, a gold or platinum weight implanted in the upper lid lets gravity help it close. Where the lower lid has become lax and no longer sits against the globe, tightening it, commonly with a lateral tarsal strip, restores the seal along the bottom. Most of these are considered once the disease has quietened, since rehabilitative surgery belongs to inactive disease. Severe exposure is the exception that does not wait.

Does treating the thyroid eye disease itself protect the cornea?

Controlling the active, inflammatory phase is what stops the anatomy getting worse, and the anatomy is what determines the exposure. Symptoms of thyroid eye disease usually last one to two years and often settle on their own, which is genuinely reassuring, but the cornea has to be defended for the whole of that time, not just at the end of it.

Disease activity control is the part of this I have spent years studying. With colleagues in Leeds I looked back over fourteen years of a tertiary oculoplastic service at activity control and steroid reduction using ciclosporin, work presented to the British Oculoplastic Surgery Society in 2026 and listed with my other research. That study is about damping down inflammation and the price of long steroid courses. It is not about the cornea. The two are linked in the clinic room every week, because the patient whose disease stays active longer is the patient whose lids are still not closing in month eighteen.

If you have thyroid eye disease and nobody has yet measured your closure, ask for it. It takes seconds. It is not the same question as whether your eyes look tired, which has a longer and much more ordinary list of causes that I go through in what actually makes eyes look tired, and it is not the same question as how far the eye protrudes. The rest of what happens inside the socket, and the conditions that share this territory, sit together under thyroid eye disease and the orbit. Appearance can be corrected later. A scarred cornea usually cannot.