She saw costumed figures walk across her living room. She was not losing her mind, she was losing her sight

There is a symptom that many people in France live with without ever mentioning it, for fear of being taken for the mentally ill. When sight declines, the visual brain, deprived of raw material, sometimes starts manufacturing its own: grids, branches, faces, silent figures crossing the room. The patient knows perfectly well that none of it is real. It has had a name since 1760, and it is one of the few symptoms in medicine that an explanation alone is enough to relieve.

GENEVA, 1760

A naturalist, his grandfather, and birds that did not exist

Charles Bonnet was an eighteenth-century Genevan naturalist and philosopher, known in his lifetime for his work on aphids and insect reproduction. In 1760 he recorded an observation that has nothing to do with entomology: that of his maternal grandfather, Charles Lullin, aged eighty-nine, operated on both eyes for cataract and left with severe sight loss.

The old man, sitting at home, sees things. Men and women coming into the room, birds, horse-drawn carriages, buildings, tapestries whose patterns unfold across the walls. The images are sharp, coloured, silent. They appear and disappear without his bidding. And above all: he knows they are not there. He describes them to his grandson with the calm curiosity of a man observing a phenomenon, not with the dread of a man in the grip of delusion.

It is precisely this point, the preserved awareness that the images are unreal, that makes the observation a landmark. Bonnet himself would later experience visual disturbances and report comparable experiences. Even so, it took until the 1930s for the Genevan neurologist Georges de Morsier to link this picture to a syndrome and give it the name of Charles Bonnet. And it took until the 1990s for it to stop being regarded as a textbook rarity.

THE CATALOGUE

What a visual cortex that receives nothing manufactures

The images patients describe are not just any images. They fall into a few families that recur from one study to another, from one country to another, with a consistency that long puzzled observers. Some are simple and geometric; others are entire scenes, peopled with figures. What they all have in common is that they are sharp, often very colourful, and that they overlay the real room without ever blending into it: the patient sees both their living room and what has been added to it.

Two details are striking when one reads patients’ descriptions. The first: the figures are frequently dressed in the style of another era, wearing hats, in costume. The second: they are silent. They do not speak, do not look at the patient, do not react to them. There are no voices, no smells, no contact. It is a strictly visual phenomenon, and that exclusivity is an important element of the diagnosis.

THE FIGURES

It is not rare. It is merely left unspoken

For two centuries Charles Bonnet syndrome was presented as a curiosity. Recent meta-analyses tell a very different story. The largest, published in 2025, pools 49 studies and 20,303 ophthalmology patients: the overall prevalence of the syndrome comes out at 10.2% (95% confidence interval: 7.2 to 14.1%). Among patients attending low vision rehabilitation, that is to say those whose impairment is most severe, it rises to 24.6%. The authors conclude that roughly one patient in six with retinal disease, glaucoma or low vision is liable to experience it.

The figures by condition point the same way. In age-related macular degeneration, a meta-analysis of 18 studies covering more than 4,300 patients finds a prevalence of 15.8%; it falls to 7.2% among patients recruited consecutively in clinic for a wet form, and climbs to 31.6% among people attending a visual rehabilitation centre. In glaucoma, a meta-analysis of 827 patients finds 2.8% across all stages, but 13.5% when visual acuity is low in both eyes, and 20.1% in rehabilitation.

These gaps are simple to read: what counts is not the disease, it is the amount of visual information lost. A prospective study in 337 patients with open-angle glaucoma did indeed show a direct correlation between visual field damage and the occurrence of hallucinations. That is no coincidence, and it already points towards the mechanism.

The most telling figure is not a prevalence

In 1996 a Dutch team published in the Lancet the study that would change how the syndrome was perceived. Of 505 visually impaired patients screened, 60 met the criteria. All of them, without exception, were fully aware that what they were seeing was not real. In 46 of them the images carried no personal meaning: they were not memories, not familiar faces, not scenes loaded with significance. Seventeen patients were genuinely distressed by them.

But the figure that counts lies elsewhere. Of the sixty patients, only sixteen had mentioned it to a doctor. And of those sixteen, only one had been given the right diagnosis. The others had gone away with no explanation, or with a wrong one. The authors note, finally, that all the patients, without exception, were relieved to learn that these images did not mean mental illness. That detail says just about everything about what silence costs.

THE MECHANISM

How we know it comes from the visual cortex, and not from the mind

The dominant hypothesis is called deafferentation. The visual cortex is an organ that has spent a whole lifetime receiving a continuous stream of information from the retina. When that stream dries up (because the macula no longer transmits, because the visual field has narrowed, because the crystalline lens no longer lets enough light through), the visual areas do not switch off. They carry on working, and part of their spontaneous activity becomes perceptible. It is the same principle as the phantom limb after an amputation, or tinnitus after hearing loss: a sensory system deprived of input produces its own output.

This hypothesis did not stay theoretical. In 1998 a London team carried out functional MRI scans in patients with the syndrome, at the precise moment they were hallucinating. The result is remarkable: the activity sits in the ventral extrastriate cortex, and (this is the decisive point) the content of the hallucination matches the specialisation of the area activated. When the patient sees colours, it is the region specialised in colour that lights up; when they see faces, it is the face region. These patients also showed increased activity in those regions, including between episodes.

Twenty years later, an Australian team supplied an electrophysiological argument. By stimulating the peripheral visual field that had remained functional in patients with macular degeneration, they measured markedly higher cortical responses in those who hallucinated than in those who did not, and than in controls with no eye disease. The visual cortex of these patients is, literally, more excitable.

One should nonetheless resist too simple a picture. A multimodal study published in 2025, combining functional MRI, electroencephalography and transcranial magnetic stimulation in 19 patients compared with 18 controls with the same eye diseases, found on the contrary a reduced activation of the primary and ventral extrastriate visual cortex, with a slowing of cortical activity. In other words: deafferentation is indeed the starting point, but exactly what it sets off in the cortex remains an open research question.

THE DIAGNOSIS

What distinguishes Charles Bonnet from a psychiatric or neurological hallucination

This is the question that troubles patients, and the one that has to be asked methodically in the consulting room. Several features point the way with good reliability.

• Awareness that the images are unreal is preserved. The patient knows that what they see is not there. They do not try to interact with the figures, do not attribute intentions to them, do not build a delusional explanation.
• The phenomenon is exclusively visual. No voices, no smells, no tactile sensation. As soon as another sensory modality is involved, the diagnosis has to be reconsidered.
• There is no confusion and no disorientation. The patient is perfectly lucid, knows where they are and what day it is, and often recounts the scene in detail.
• There is documented sight loss. That is the obligatory context: retina, optic nerve, cornea or transparent media. Without visual impairment, it is not Charles Bonnet syndrome.
• The images carry no personal meaning. In the Dutch study this was the case in more than three quarters of patients: strangers, patterns, anonymous scenes.

Conversely, certain signs mean the enquiry has to be widened. Hallucinations accompanied by voices, by unshakeable convictions or by a sense of persecution point towards a psychiatric disorder. Visual hallucinations associated with fluctuating alertness, with slowness or motor rigidity, or with disturbed behaviour during sleep, should raise the possibility of a neurodegenerative disease, Lewy body disease in particular, in which awareness that the images are unreal is often lost. Sudden onset with confusion and disorientation suggests acute delirium, which is an emergency in its own right. Finally, hallucinations associated with unusual headache or with a neurological deficit warrant an opinion without delay.

One honest qualification has to be added: Charles Bonnet syndrome and early cognitive decline can coexist in the same older person. The presence of one does not rule out the other. That is precisely why the symptom deserves to be named and assessed, rather than hidden.

IN PRACTICE

What sets the visions off, and what makes them stop

The patients in the Dutch study described two predisposing circumstances that have recurred throughout the literature ever since: sensory deprivation and a low level of arousal. In concrete terms that means a dimly lit room, the end of the day, sitting still, tiredness, boredom, the absence of conversation. The images rarely appear in a busy street at midday; they appear in an armchair, at dusk, in silence.

This observation has an immediate practical consequence, because it can be turned into advice. The review articles devoted to management list simple manoeuvres, tested in small numbers but easy to try, and with nothing to lose:

• Change the lighting. Switch on a lamp when the vision appears in half-light, or, conversely, tone down a dazzling light.
• Move the eyes. Looking alternately left and right for a few seconds, without moving the head, interrupts the episode in some patients.
• Blink, or close the eyes and open them again. Some report that the image dissolves; others that it moves.
• Change activity or change room. Stand up, walk, switch the radio on, call someone. Coming out of a state of passive wakefulness often works.
• Stare straight at the image. Counter-intuitive, but reported as effective by some patients.

None of these manoeuvres is a treatment in the strict sense. They simply give the patient a hold on a phenomenon that would otherwise be entirely imposed on them; and that is already considerable.

THE CONSULTATION

What an ophthalmologist can actually offer

There is no medicine whose effectiveness is established in this indication. The published trials involve very small numbers, with no control group and contradictory results. That does not mean there is nothing to be done: on the contrary, management rests on three concrete levers.

The first lever is the explanation itself. The authors of the 2025 meta-analysis put it bluntly: the information given to the patient brings them relief. Knowing that this phenomenon has a name, that it affects roughly one patient in six among those with retinal disease, glaucoma or low vision, that it follows from the workings of the visual cortex and not from a mental disorder, radically changes the way it is experienced. Many patients then stop dreading it, and some simply stop hiding it from those around them.

The second lever: treating what can be treated

Since the starting point is the loss of visual information, anything that restores that information deserves to be considered. An operable cataract is the most favourable case: surgery gives the cortex back a bright, contrasted stream of light that it was no longer receiving. Wet AMD calls for a treatment whose regularity determines the outcome; an atrophic form calls for monitoring and rehabilitation. Advanced glaucoma justifies securing the intraocular pressure in order to preserve what remains of the visual field. An unsuitable optical correction, a pterygium, a corneal opacity, severe dryness: each of these removes signal, and each can be treated. Cataract surgery, like any operation, carries risks that must be set out at the pre-operative consultation, and the indication for it is assessed individually.

The third lever is low vision rehabilitation. Optical and electronic aids, adapting the lighting at home, enhanced contrast, orthoptics, psychological support where the impact is significant. The apparent paradox (the highest prevalences are measured in rehabilitation centres) has a simple explanation: these centres receive the most severe impairments, and they are also the places where somebody thinks to ask the question.

One thing remains hard to predict: the course of the condition. In many patients the hallucinations fade or disappear over a few months to a few years; in others they persist, simply becoming familiar. The literature does not allow a duration to be announced to any given patient, and it is more honest to say so than to improvise a prognosis.

KEY POINTS

What to remember

• Seeing patterns, figures or scenes while sight is declining is a phenomenon described since 1760 and documented by dozens of studies. It is not a sign of mental illness.
• Its frequency is of the order of 10% in general ophthalmology, and about one patient in four in low vision rehabilitation.
• The patient generally retains the awareness that the images are unreal, and the phenomenon is exclusively visual: no voices, no smells, no bodily sensations.
• The accepted mechanism is deafferentation of the visual cortex, supported by functional imaging and electrophysiology, even if not everything about it has been elucidated.
• There is no validated medicine, but explanation, treatment of the visual cause where that is possible, and rehabilitation genuinely change how it is lived with.
• Mentioning it to your ophthalmologist is worthwhile: it is what makes it possible to tell this syndrome apart from the situations that do call for different management.

FREQUENTLY ASKED QUESTIONS

Frequently asked questions

Scientific sources

1. Visual hallucinations in psychologically normal people: Charles Bonnet’s syndrome. Lancet. 1996. PMID: 8622335
2. The Charles Bonnet syndrome: a large prospective study in The Netherlands. Br J Psychiatry. 1995. PMID: 7728372
3. The prevalence of Charles-Bonnet syndrome in ophthalmic patients: a systematic review and meta-analysis. Brain Res Bull. 2025. PMID: 40049460
4. Prevalence of Charles Bonnet syndrome in patients with age-related macular degeneration: systematic review and meta-analysis. Acta Ophthalmol. 2020. PMID: 31654492
5. Prevalence of Charles Bonnet syndrome in patients with glaucoma: a systematic review with meta-analyses. Acta Ophthalmol. 2021. PMID: 32749787
6. Charles Bonnet syndrome in patients with open-angle glaucoma: prevalence and correlation to visual field loss. Ophthalmol Glaucoma. 2022. PMID: 34688955
7. The anatomy of conscious vision: an fMRI study of visual hallucinations. Nat Neurosci. 1998. PMID: 10196592
8. Stimulus-driven cortical hyperexcitability in individuals with Charles Bonnet hallucinations. Curr Biol. 2018. PMID: 30415703
9. Visual cortical activity in Charles Bonnet syndrome: testing the deafferentation hypothesis. J Neurol. 2025. PMID: 39932561
10. Investigation of structural brain changes in Charles Bonnet syndrome. Neuroimage Clin. 2022. PMID: 35576854
11. Rehabilitation in Charles Bonnet syndrome: a review of treatment options. Clin Exp Optom. 2004. PMID: 15186205
12. Charles Bonnet syndrome, insight and cognitive impairment. J Am Geriatr Soc. 1997. PMID: 9215348

Further reading

• AMD: detection, forms and management
• Glaucoma: detection and visual field monitoring
• Cataract symptoms: when to seek advice
• Dr Moïse Tourabaly — Ophthalmologist

Cachan practice · Tel. +33 1 45 47 08 11

Medical disclaimer

This article is for information only. A personalised ophthalmological opinion remains essential for any treatment decision.

The patterns shown in the interactive module are graphic reconstructions produced from descriptions published in the medical literature. They do not reproduce what any individual person sees and are neither a test nor a screening tool: no conclusion about your own vision can be drawn from them. The prevalence figures quoted come from the meta-analyses referenced above and depend on the population studied, which explains their spread. Any occurrence of visual hallucinations, particularly when associated with other symptoms, warrants a medical opinion, which alone can determine the cause.