In the present study, a CBS prevalence of 26% was found in patients with a VA ≥ 0.5 logMAR. This confirms the range of previously reported prevalences of CBS in patients with low vision (6–34%). A recent meta-analysis found a prevalence of 19.7% among low-vision patients [26]. Furthermore, the prevalence of CBS in low-vision patients is comparable to other populations with various ophthalmic conditions, such as glaucoma patients (prevalence = 23%, N = 141 [12]), patients with retinal diseases (prevalence = 40%, N = 53 [8]; prevalence = 38%, N = 72 [27], prevalence = 39%, N = 1254 [28]) and in vision rehabilitation centres (prevalence = 35%, N = 225 [29]; prevalence = 24%, N = 50 [30]; prevalence = 19%, N = 2565 [15]). These findings suggest that low vision and ocular pathology are in fact risk factors for the development, as CBS occurred in all groups. Another factor might be the residual visual field. A previous study reported a correlation between VH and visual field recovery in patients undergoing vision restoration therapy (VRT) [31]. The study found that patients with large residual visual areas and diffuse visual field defects were more likely to experience VH during training stimuli. To date, only one study has investigated the occurrence of CBS in relation to the percentage of central and peripheral vision loss [32], and found a correlation between visual field and occurence of CBS. Reasons for different prevalences of existing studies could be due to both the varying definitions of CBS and different inclusion criteria. Sometimes researchers require their patients to have only complex hallucinations [9, 17] or exclude patients with hallucinations in other modalities [19, 20]. However, in a study investigating the association between visual and auditory hallucinations, no particularly high prevalence was found [18]. Furthermore, it might be possible that patients’ reluctance to talk about hallucinations, even when asked, varies from country to country due to cultural differences. For example, some patients may not want to report them for fear of being labelled as “mentally ill”, a genuine concern as some case reports show [23, 33, 34]. In one study, about one in ten patients attributed the hallucinations to mental illness [28].

Statistically, women were not significantly more often affected from CBS than men. However, our data suggest a preponderance of women among CBS patients, as also found in other studies [7, 8, 15, 23, 33, 35,36,37,38,39,40,41,42,43,44]. Mostly, it is assumed that the difference in sex distribution results from the fact that women tend to communicate more deliberately about unfamiliar events [15]. In a case report, Fernandes et al. observed a female patient who started to hallucinate only after the administration of oestrogens for osteoporosis therapy [45]. After discontinuation of the treatment, the VH disappeared. Maybe oestrogen could be a mediator in relation to the sex distribution in CBS. However, further studies are needed to confirm or disprove this theory.

Often, insight into the unreality of images is not gained immediately. 46.2% of patients did not initially recognise the images as unreal. Insight in the unreality of the hallucinations is a key feature that is often required for CBS diagnosis. De Morsier, who first described the syndrome in 1967, did not consider insight to be mandatory, as it can be subject to fluctuations [2]. Other commonly used diagnostic criteria such as those of Damas-Mora et al. [4] and Rabins [3] insist that insight, at least partial insight, is crucial for the diagnosis of CBS. In this context, it is important to note that the present study found that a considerable proportion of patients did not initially recognize the hallucinations as such, also recognised in the literature [9]. Yet, full insight is often an inclusion criterion, which may result in patients with delayed insight in the VH being excluded from some studies. There are few studies examining the onset for VH in relation to the onset of insight in them. Typically, delayed but then quickly gained insight was observed [11, 33, 34, 43, 46, 47].

Patterns, or so-called “simple” hallucinations, occurred just as frequently as other types of images such as people, body parts, or faces [18, 20, 33, 48,49,50,51]. In line with other studies [7, 40], the most frequent images were animals. The occurrence of simple hallucinations sometimes constitutes an exclusion of CBS. A variety of images are described as simple hallucinations, including flashes, sparks, bugs, or colour fields, snowflakes, mist, geometric shapes and patterns [33, 49]. However, simple hallucinations have been reported in up to 23% of patients who saw at least one other complex image [49]. The requirement that as least one complex image occurs alongside other, simpler hallucinations might be a useful tool to make an accurate differential diagnosis. Nevertheless, other studies have also observed the familiarity of simple hallucinations in CBS patients [11, 27, 28, 33, 40, 42, 48], with, for example, up to 63% of patients seeing patterns as a form of simple hallucinations in a 2014 survey by Cox and Ffytche [28].

Visual hallucinations occurred more frequently during the day (46.2%) and in bright surroundings (46.2%). In most studies, evening/night time [8, 20, 48, 52] and poor or dim lighting [7, 20, 40, 42, 48] were mentioned as favourable circumstances for the appearance of VH. In addition, it was observed in one case that improved lighting resolved the hallucinations [53], and in another study about one in five patients recognised the lighting intensity as potential trigger for VH [11]. These findings were not confirmed by the present study. However, it does not directly contradict the literature. A non-negligible proportion of patients describes bright or dazzling lights to trigger CBS hallucinations [7, 27, 40, 42]. Moreover, hallucinations may occur without any circadian affinity [48] or independently of ambient lighting [27]. Painter et al. showed that early visual cortex is hyperexcitable in CBS patients [54], suggesting that in an hyperexcitable cortex any stimulus, whether a poor visual input in the dark or a blurry but bright visual input in well-illuminated surroundings, can lead to overactivation of neurons. The brain would thereby create visual percepts with no correlation in the physical world, i.e. visual hallucinations. Although it may seem contradictory, these two clinical manifestations of CBS could be due to the same pathophysiology.

In addition, all patients experienced the hallucinations only with their eyes open. The hallucinations were generally (61.5%) not moving with the eyes. Literature shows that hallucinations with eyes open are quite common. In various studies, a large proportion of patients reported that their hallucinations occurred with their eyes open [11, 20, 48, 49, 52], spanning from 37.5[52]–97.9% [48]. Sometimes patients are also able to resolve their hallucinations by closing their eyes [33, 55, 56] suggesting that there is a close interaction between the brain and the eye in the development of the VH. The images do not arise spontaneously, but are triggered by an incomplete visual impression due to vision loss or visual field loss. The findings presented thus support the hypothesis of sensory deprivation (deafferentation) on the development of CBS hallucinations [57], emphasizing the importance of ocular pathology in the etiology of CBS. According to theory, ocular pathology with subsequently reduced visual input leads to spontaneous discharges in the visual cortex (hyperexcitability). As mentioned before, individuals with CBS do indeed have a hyperexcitable visual cortex [54]. The interaction between the diseased visual pathway and the brain is therefore crucial for the development of CBS. However, it should be noted that patients sometimes hallucinate with their eyes closed [58]. This interaction is also reflected in the fact that the visual hallucinations often do not move with the eyes, but generally stay in a certain place [20, 33]. However, this characteristic has only been investigated in a few studies. Sometimes the images themselves can also move across the visual field, such as marching soldiers [59]. Hallucinations seem to fit easily into the patient’s current spatial perception, implying again a close interaction between brain and visual pathway is underlying the aetiology of CBS and that spatial awareness is most likely unaffected by CBS.

38.5% of CBS sufferers did not communicate about their hallucinations. Most patients (76.9%) did not consult any physician. It has been observed in literature that many patients feel unwell because of their hallucinations, at least to some extent [20, 28, 33, 40, 48, 49]. Hallucinations can have a negative impact on the patients ‘ lives [20, 40], e.g. by interfering with daily activities [