Charles Bonnet Syndrome

Abstract

Charles Bonnet syndrome (CBS) is the occurrence of complex visual hallucinations in people who have lost a substantial part of their sight but who are psychologically well, fully aware that what they see is not real, and free of any psychiatric disorder. The hallucinations range from simple patterns and lattices to vivid, detailed images of people, animals, faces, and landscapes, and they arise not from disease of the mind but from disease of the eye or visual pathway. The condition is understood as a release phenomenon: when the visual cortex is deprived of its normal input, it becomes spontaneously overactive and generates its own imagery. CBS matters to cognitive psychology because it demonstrates that perception is not a passive reception of the world but an active construction the brain continues even in the absence of a stimulus.

Keywords: Charles Bonnet syndrome, visual hallucinations, deafferentation, release phenomenon, vision loss

Charles Bonnet syndrome occupies an unusual place in the study of perception because it separates two things ordinarily bound together: seeing and having something to see. A person with advanced macular degeneration, glaucoma, or diabetic retinopathy may have almost no usable vision, yet report seeing brickwork patterns spread across a wall, small costumed figures walking across the floor, or the face of a stranger looking back at them — images so detailed and so external that they must be reassured the experiences are a known consequence of eye disease, not a sign of madness (Menon, Rahman, Menon, & Dutton, 2003). The defining feature is that insight is preserved: the person knows the visions are not real, which is precisely what distinguishes CBS from a psychotic hallucination.

The account below moves from what the syndrome is, through the characteristic range of its hallucinations, to the deafferentation or “release” hypothesis that explains why a failing eye should produce vivid imagery in the mind, then to the cortical mechanisms that imaging and stimulation studies have revealed, and finally to how the condition is diagnosed and managed. The recurring theme is that these hallucinations are not a psychiatric symptom but a signature of how the visual brain behaves when it is starved of input — a window onto the constructive, self-generating nature of normal perception.

What Charles Bonnet Syndrome Is

Charles Bonnet syndrome is the occurrence of complex visual hallucinations in a person with significant vision loss who is otherwise mentally well and who recognises the hallucinations as unreal. The syndrome is named for the Swiss naturalist and philosopher Charles Bonnet, who in 1760 described the case of his grandfather, Charles Lullin, who after cataract surgery began to see vivid figures, birds, and buildings that he knew did not exist (Damas-Mora, Skelton-Robinson, & Jenner, 1982). The essential elements have not changed since: intact insight, an absence of psychiatric illness, and hallucinations in the visual modality alone, occurring against a background of damaged vision.

Three features together define the condition and separate it from other causes of hallucination. First, the person has acquired vision loss — from age-related macular degeneration, glaucoma, diabetic retinopathy, or any other disease of the eye or visual pathway. Second, the hallucinations are purely visual: there are no voices, smells, or tactile experiences, which distinguishes CBS from the multimodal hallucinations of psychosis or delirium. Third, and most importantly, insight is preserved — the person understands that the visions are a product of their own visual system, not real events in the world, even when the images are strikingly lifelike (Teunisse et al., 1996).

That preserved insight is the diagnostic linchpin. It is what allows CBS to be recognised as a benign consequence of eye disease rather than a psychiatric emergency, and it is why the single most effective intervention is often simply telling the patient what the syndrome is. Many people conceal their hallucinations for years, fearing that to admit them is to admit to losing their mind; learning that the experience has a name, a mechanism, and no bearing on their sanity is itself therapeutic (Menon et al., 2003). CBS is, in this sense, as much a problem of clinical recognition as of underlying pathology.

The Phenomenology of the Hallucinations

The hallucinations of Charles Bonnet syndrome fall along a spectrum of complexity. At the simple end are elementary hallucinations: flashes of light, coloured dots, and geometric forms such as grids, lattices, and tessellated patterns, often called photopsias when they are mere points or lines. At the complex end are fully formed hallucinations: recognisable people, faces, animals, plants, vehicles, and extended scenes, sometimes in period costume, sometimes miniature, frequently in vivid colour and fine detail (ffytche et al., 1998). A recurring and distinctive motif is the appearance of small figures — so-called Lilliputian hallucinations — and of repeating patterns of brickwork, fencing, or mosaic.

Certain properties recur across cases and are clinically useful. The images are typically not under voluntary control: they appear and vanish unbidden, may last seconds or hours, and are often triggered by low illumination, fatigue, or inactivity. They are usually experienced as external and located in the world rather than in the mind's eye, and they do not interact with the person or respond to their actions (ffytche, 2009). Emotionally they are most often neutral or even mildly pleasant, though a minority of patients find them frightening or distressing — particularly when the content is disturbing rather than benign, a variant sometimes called negative-outcome CBS that carries a real psychological toll (Cox & ffytche, 2014).

Crucially, the content of the hallucination tends to reflect the type and site of visual loss, and the complexity of a hallucination is not random. This lawful structure — simple percepts from early visual disturbance, formed percepts implicating specialised higher visual areas — is what first suggested that CBS is generated within the visual system itself, and it is the observation the neural accounts below must explain (Santhouse, Howard, & ffytche, 2000).

The Deafferentation, or Release, Hypothesis

Why should a failing eye produce more visual experience rather than less? The leading answer is the deafferentation, or release, hypothesis. Under normal conditions, the visual cortex is continuously driven and constrained by the stream of signals arriving from the retina; that input both supplies content and, through ongoing activity, holds the cortex's own spontaneous tendencies in check. When disease interrupts the input — deafferenting the cortex — the cortex is not merely quieter; it becomes disinhibited and hyperexcitable, and begins to generate activity of its own. That endogenous activity, interpreted by the rest of the visual system in the only terms it has, is experienced as a percept (Burke, 2002).

The analogy often drawn is with phantom-limb sensations, in which the deafferented cortex that once served an amputated limb continues to generate experiences referred to the missing part. On this view, a Charles Bonnet hallucination is a phantom vision: the perceptual system filling a void left by lost input with its own construction (ffytche, 2005). The account elegantly explains the syndrome's central paradox — that hallucinations often begin or worsen as sight deteriorates, and may actually recede if some vision is restored — because it is the loss of input, not any positive lesion, that releases the activity.

The release framework also predicts the lawful mapping between the site of visual loss and the character of the hallucination. Because the visual cortex is functionally parcellated — distinct regions for colour, form, motion, and faces — deafferentation of a particular specialised region should release activity bearing that region's signature, yielding coloured patterns from one area and formed faces from another. This is exactly the correspondence that imaging studies went on to confirm, and it is why the release hypothesis remains the organising idea of the field even as its cellular details are refined (Carpenter, Jolly, & Bridge, 2019).

The Neural Basis

The decisive evidence for the release account came from functional imaging that caught the hallucinating brain in the act. In a landmark study, activity in the ventral visual cortex rose precisely when patients reported hallucinations, and — the key finding — the location of that activity corresponded to the content of the experience: hallucinations of colour engaged colour-selective cortex, hallucinations of faces engaged the fusiform face region, and hallucinations of printed text and objects engaged the corresponding specialised areas (ffytche et al., 1998). The content-specificity was strong evidence that the hallucination is generated by the same cortical machinery that supports normal perception of that content, activated from within rather than by a stimulus.

This anatomical logic was extended by comparing across the visual hallucinatory syndromes. Mapping the phenomenology of hallucinations onto the functional architecture of the visual brain showed a consistent correspondence between the type of hallucination and the visual area implicated, unifying CBS with hallucinations arising from other causes under a common anatomical scheme and reinforcing that these experiences are organised by the structure of the visual system itself (Santhouse et al., 2000). More broadly, CBS takes its place within a wider taxonomy of disorders of visual perception, alongside other conditions in which the perceptual construction of the world goes awry (ffytche, Blom, & Catani, 2010).

The cortical mechanism has since been probed more directly. Behavioural and physiological work indicates that people who experience Charles Bonnet hallucinations show heightened, stimulus-driven excitability of the visual cortex compared with visually impaired people who do not hallucinate — a measurable predisposition of the deprived cortex toward the very hyperactivity the release hypothesis posits (Painter, Dwyer, Kamke, & Mattingley, 2018). The emerging picture is one in which vision loss provides the deafferentation and an intrinsically hyperexcitable visual cortex provides the tendency to fill the resulting void, so that the hallucination is the joint product of reduced input and released cortical activity (ffytche, 2009).

Epidemiology and Risk Factors

Charles Bonnet syndrome is far more common than its rate of diagnosis suggests. In populations of people with significant visual impairment, prevalence estimates commonly fall in the range of roughly 10–15%, with the classic large study of visually handicapped patients placing it near 11% for complex hallucinations, and higher figures reported when simple hallucinations are included or when patients are asked directly (Teunisse et al., 1996). Because the syndrome is driven by vision loss, its prevalence rises with the severity of visual impairment and with the eye diseases that cause it, and it is therefore most often seen in older adults, in whom age-related macular degeneration and other sight-threatening conditions are concentrated (ffytche, 2009).

The dominant risk factor is the degree of visual acuity loss: the more sight is lost, the more likely hallucinations become, consistent with the deafferentation account. Bilateral disease, reduced contrast sensitivity, and social isolation or sensory understimulation are also associated with the syndrome, and the hallucinations characteristically emerge or intensify in conditions of low sensory input — dim light, eyes closed, or during periods of inactivity (Pang, 2016). Cognitive status is generally normal, which is central to the definition, though the relationship between CBS and later cognitive decline is a question of ongoing study.

A crucial epidemiological fact is under-reporting. Surveys consistently find that most patients do not spontaneously mention their hallucinations, and that a large fraction had never told anyone before being asked, chiefly for fear of being judged mentally ill (Schadlu et al., 2009). Prevalence figures therefore depend heavily on how the question is posed: passive waiting for disclosure catches only a minority of true cases, while direct, routine questioning in low-vision settings uncovers far more. This gap between how common the syndrome is and how often it is recognised is the central practical problem the syndrome presents.

Diagnosis and Differential Diagnosis

Diagnosis of Charles Bonnet syndrome is clinical and rests on a characteristic pattern rather than any test. The core criteria are formed or simple visual hallucinations occurring in a person with visual impairment, in the presence of intact insight, in the absence of hallucinations in other sensory modalities, and without evidence of a psychiatric disorder, delirium, or a neurological condition that would better account for the experiences. Because the field long lacked agreed criteria, recent work has sought to standardise the definition and the diagnostic threshold, which matters both for individual diagnosis and for comparing prevalence across studies (Hamedani & Pelak, 2019).

The differential diagnosis is where the distinctive features earn their keep. Unlike the hallucinations of a psychotic illness, those of CBS are visual only, are recognised as unreal, and carry no delusional interpretation or accompanying thought disorder. Unlike the hallucinations of delirium, they occur in a clear, un-clouded sensorium without fluctuating attention or confusion. And unlike the visual hallucinations of dementia with Lewy bodies or Parkinson's disease, classic CBS occurs without the motor signs, cognitive decline, or loss of insight that characterise those conditions (ffytche et al., 2010). The single most useful discriminator across all of these is preserved insight coupled with an isolated visual modality against a background of known eye disease.

Assessment therefore combines a careful history of the hallucinations' content, timing, and triggers with documentation of the visual impairment and a screen for the features that would point elsewhere — other-modality hallucinations, clouding of consciousness, delusions, or neurological signs. Where the picture is atypical, further investigation is warranted, but in the typical case of an older adult with sight loss and benign, insight-retained visual imagery, the diagnosis is secure and the most important step is explanation and reassurance (Pang, 2016).

Management

The cornerstone of management is explanation. Because so much of the distress in Charles Bonnet syndrome comes from the fear that the hallucinations signify mental illness, simply naming the condition, explaining its mechanism as a consequence of eye disease, and reassuring the patient that it does not indicate psychiatric disorder is often the most valuable intervention and, for many, sufficient in itself (Schadlu et al., 2009). This reframing converts a frightening, secret experience into an understood and expected one, and it is the reason routine enquiry in low-vision care is so important: recognition is treatment.

Non-pharmacological strategies aim to modulate the sensory conditions that provoke the hallucinations. Because the imagery tends to emerge under low input, patients are often advised to improve lighting, to blink or move the eyes or look away when a hallucination appears, to increase social and cognitive stimulation, and, where possible, to optimise their remaining vision through low-vision aids and treatment of the underlying eye disease (Menon et al., 2003). Where a hallucination can be interrupted by a change in gaze or illumination, patients can be taught these manoeuvres as a form of self-management.

Pharmacological treatment is reserved for the minority whose hallucinations are frequent, distressing, or disabling, and no drug is established as reliably effective; the evidence base is limited to small series and case reports, and any medication must be weighed against its side effects, particularly in older adults (Pang, 2016). The overall management philosophy is therefore conservative and staged: recognise and explain first, adjust the sensory environment second, and consider medication only when the burden justifies it.

Worked Example

Consider how under-reporting shapes what a clinic actually sees. Suppose a low-vision clinic assesses 400 patients with significant visual impairment in a year, and that the prevalence of complex Charles Bonnet hallucinations in such a population is about 11% (Teunisse et al., 1996). The expected number of true cases is:

expected cases = 400 × 0.11 = 44

So roughly 44 of these patients are experiencing CBS. But suppose — consistent with the repeated finding that most patients never volunteer their hallucinations — that only about one in four spontaneously disclose them without being asked. The number the clinic detects by passively waiting is then:

detected (passive) = 44 × 0.25 = 11

The clinic sees 11 cases and misses 44 − 11 = 33. Now suppose the clinic introduces a single direct question about visual hallucinations into every low-vision assessment, and that routine questioning elicits disclosure from around 90% of true cases:

detected (direct question) = 44 × 0.90 ≈ 40

The same underlying prevalence now yields about 40 recognised cases instead of 11 — nearly a fourfold increase — simply by asking. The example makes the epidemiological point concrete: the gap between how common CBS is and how often it is diagnosed is not a gap in prevalence but a gap in enquiry, which is why direct questioning is the single most effective diagnostic act in the syndrome (Schadlu et al., 2009). (The 25% and 90% disclosure figures are illustrative; the 11% prevalence is drawn from the cited study.)

Discussion

Charles Bonnet syndrome matters to cognitive psychology because it makes vivid a claim that is otherwise abstract: that perception is an active construction, not a passive recording. If seeing were merely the reception of retinal signals, then losing those signals should produce only darkness. Instead it can produce a surplus of vision — detailed, coloured, formed imagery generated by a cortex that has lost its input but not its inclination to build a visual world (Burke, 2002). The syndrome thus supplies direct human evidence that the visual brain is a generative system continuously predicting and constructing, and that sensory input serves as much to constrain that construction as to create it.

The content-specificity revealed by imaging deepens the point. That hallucinations of faces engage face cortex and hallucinations of colour engage colour cortex shows that the brain's functional parcellation is not merely a map of how incoming stimuli are analysed but a map of how experience is manufactured, whether from without or from within (ffytche et al., 1998). The same specialised region can be driven by a real face or released into activity by deafferentation, and in both cases the result is the experience of a face. CBS therefore blurs the line between perception and imagery in a theoretically instructive way, and it connects to broader questions about how the brain distinguishes internally generated activity from externally caused activity — a distinction that, in CBS, remains intact enough to preserve insight.

Open questions remain about why some visually impaired people develop the syndrome and others with comparable sight loss do not. The finding of heightened cortical excitability in those who hallucinate suggests that a predisposition of the cortex, and not deafferentation alone, is required, pointing toward an interaction between input loss and intrinsic cortical properties (Painter et al., 2018). Resolving this interaction bears directly on prediction and, eventually, on targeted treatment, and it keeps CBS at the intersection of clinical ophthalmology and the cognitive neuroscience of perception.

Current Directions

Contemporary research is moving on several fronts. One is the drive toward standardised diagnostic criteria: because the field has long operated without an agreed definition or threshold, a systematic effort to codify the criteria is a prerequisite for consistent diagnosis, reliable prevalence estimates, and comparable clinical trials (Hamedani & Pelak, 2019). Settling what does and does not count as CBS is the unglamorous but foundational work on which the rest depends.

A second front is mechanistic. Building on the release hypothesis, current work seeks to specify the cellular and network changes that make a deafferented visual cortex hyperexcitable, integrating the classical account with contemporary models of cortical gain, disinhibition, and the balance between feedforward input and internally generated prediction (Carpenter et al., 2019). The demonstration of stimulus-driven hyperexcitability in hallucinators is a step toward a physiological marker that could identify who is at risk before hallucinations begin (Painter et al., 2018).

A third front broadens the clinical picture beyond the older adults in whom CBS is most familiar. Recent work documents the syndrome in children and young adults with sight loss, showing that it is neither confined to the elderly nor dependent on age-related brain change, and underscoring that clinicians across the age range should ask about hallucinations in anyone with significant visual impairment (Jones & Moosajee, 2021). Together these directions are shifting CBS from a clinical curiosity toward a well-characterised, criteria-defined, mechanistically grounded consequence of vision loss.

Common Misconceptions

“Visual hallucinations mean the person is developing a mental illness.”

Not in Charles Bonnet syndrome. The hallucinations arise from eye disease, not psychiatric disease; insight is preserved, the experiences are visual only, and there is no delusion or thought disorder. Mistaking CBS for psychosis is the error the diagnosis exists to prevent (Teunisse et al., 1996).

“It is a rare, unusual condition.”

It is common in people with significant vision loss — on the order of one in ten — but greatly under-diagnosed because patients rarely disclose the hallucinations and clinicians rarely ask (Schadlu et al., 2009).

“The hallucinations are a sign the eye disease is spreading to the brain.”

They are not. They are a release phenomenon caused by the brain being deprived of visual input; the visual cortex generates activity of its own precisely because it has lost its normal signal, not because disease has invaded it (Burke, 2002).

“Nothing can be done about it.”

Explanation and reassurance alone substantially reduce distress for many patients, and simple measures — better lighting, changes of gaze, increased stimulation, optimising remaining vision — can lessen the hallucinations. Recognition is itself an effective intervention (Menon et al., 2003).

Glossary

Charles Bonnet syndrome.

The occurrence of complex visual hallucinations in a person with acquired vision loss who retains insight that the images are unreal and who has no psychiatric disorder.

Complex visual hallucination.

A hallucination of a fully formed, recognisable image — a person, face, animal, or scene — as distinct from an elementary flash or pattern; the hallmark content of Charles Bonnet syndrome.

Contrast sensitivity.

The ability to distinguish an object from its background across differences in luminance; its reduction is one of the visual deficits associated with a raised risk of the syndrome.

Cortical hyperexcitability.

An increased tendency of cortical neurons to fire; in CBS, the deprived visual cortex becomes hyperexcitable and generates the spontaneous activity experienced as hallucination.

Deafferentation.

The loss of sensory input to a region of cortex through damage to the pathway that normally supplies it; in CBS, the loss of visual input to visual cortex.

Formed hallucination.

A hallucination of a recognisable object — a person, face, animal, or scene — as opposed to a simple pattern or flash.

Insight.

The person's awareness that the hallucinations are not real; its preservation is the defining feature separating CBS from psychotic hallucination.

Lilliputian hallucination.

A hallucination of miniature people or figures, a distinctive motif frequently reported in Charles Bonnet syndrome.

Macular degeneration.

Progressive disease of the central retina and a leading cause of the acquired vision loss that predisposes to Charles Bonnet syndrome, especially in older adults.

Negative-outcome Charles Bonnet syndrome.

A variant in which the hallucinated content is disturbing rather than neutral or benign, carrying a genuine psychological toll for the patient.

Phantom vision.

The analogy, drawn from phantom-limb sensation, that a CBS hallucination is the perceptual system filling the void left by lost visual input with its own construction.

Photopsia.

A simple, elementary visual hallucination such as a flash of light, spark, or coloured dot, at the low-complexity end of the CBS spectrum.

Release phenomenon.

Activity that emerges when a normally constraining input is removed; the framework in which CBS hallucinations are understood as released spontaneous cortical activity.

Ventral visual cortex.

The occipitotemporal stream of specialised visual areas — for colour, form, and faces — whose released activity determines the content of a Charles Bonnet hallucination.

Visual acuity.

The sharpness of central vision; its degree of loss is the dominant risk factor for developing Charles Bonnet syndrome.

Key Researchers

Jan Dirk Blom

(living). Psychiatrist at Leiden University and the Parnassia Psychiatric Institute, The Hague, and author of A Dictionary of Hallucinations; his work situates Charles Bonnet syndrome within a comprehensive taxonomy of disorders of visual perception (ffytche, Blom, & Catani, 2010). ORCID

Charles Bonnet

(1720–1793). Swiss naturalist and philosopher who in 1760 recorded his grandfather Charles Lullin's vivid visual hallucinations following vision loss — the first description of the phenomenon that now bears his name (Damas-Mora, Skelton-Robinson, & Jenner, 1982). Wikipedia

Dominic H. ffytche

(living). Old-age psychiatrist and neuroscientist at the Institute of Psychiatry, Psychology & Neuroscience, King's College London, and the leading contemporary authority on visual hallucinations; his functional-imaging work established the content-specific cortical basis of Charles Bonnet hallucinations (ffytche et al., 1998; ffytche, 2005). ORCID

Robert J. Howard

(living). Professor of old-age psychiatry at University College London, co-author of the foundational neuroimaging and anatomical work mapping the visual hallucinatory syndromes onto the functional architecture of the visual brain (ffytche et al., 1998; Santhouse, Howard, & ffytche, 2000). ORCID

Jason B. Mattingley

(living). Cognitive neuroscientist at the Queensland Brain Institute, University of Queensland, whose work demonstrated heightened stimulus-driven cortical excitability in people who experience Charles Bonnet hallucinations, giving the release hypothesis a measurable physiological correlate (Painter, Dwyer, Kamke, & Mattingley, 2018). ORCID

Mariya Moosajee

(living). Ophthalmologist and geneticist at the UCL Institute of Ophthalmology and Moorfields Eye Hospital, whose research extended the recognition of Charles Bonnet syndrome to children and young adults with sight loss (Jones & Moosajee, 2021). ORCID

Frequently Asked Questions

What is Charles Bonnet syndrome?

Charles Bonnet syndrome is the experience of complex visual hallucinations in people who have lost a significant amount of their sight but who are mentally well and understand that the images are not real. The hallucinations are caused by the eye disease and the resulting loss of visual input to the brain, not by any psychiatric illness.

Does having these hallucinations mean I am losing my mind?

No. Preserved insight — knowing the visions are not real — is a defining feature of the syndrome, and the hallucinations are purely visual, with no voices or other sensations and no delusions. They are a recognised, benign consequence of vision loss, and many people are enormously relieved to learn the experience has a name and a harmless explanation.

What do the hallucinations look like?

They range from simple flashes, dots, and geometric patterns such as grids and brickwork to fully formed images of people, faces, animals, plants, and detailed scenes, often in vivid colour. Miniature figures are especially characteristic. The images usually appear and disappear on their own and are seen as located out in the world.

Why does losing vision cause hallucinations?

Because the visual part of the brain, deprived of its normal input, becomes overactive and generates activity of its own — a “release” phenomenon. That spontaneous activity is experienced as imagery, much as an amputated limb can still be felt. It is the loss of input, not damage to the brain, that releases the hallucinations.

How common is it?

It is common among people with serious sight loss, affecting roughly one in ten or more, but it is greatly under-diagnosed because patients often hide the hallucinations for fear of being thought mentally ill, and clinicians often do not ask about them.

How is it diagnosed?

It is a clinical diagnosis based on the pattern: visual-only hallucinations, preserved insight, an absence of psychiatric or neurological disease that would better explain them, and a background of eye disease or vision loss. There is no specific test; the key is recognising the characteristic picture and ruling out other causes such as delirium or psychosis.

Can it be treated?

The most important step is explanation and reassurance, which by itself greatly reduces distress for many people. Simple measures such as improving lighting, changing where one looks when a hallucination appears, and increasing activity and social contact can help. Medication is reserved for the minority with severe, distressing hallucinations and is not reliably effective.

Will the hallucinations ever go away?

They often fluctuate and may lessen over time, particularly if the underlying vision stabilises or improves, or as the brain adjusts. For many people they become less frequent or less troubling once understood, though the course varies from person to person.

References