Lilliputian hallucinations concern hallucinated human, animal or fantasy entities of minute size

Leroy’s elusive little people: A systematic review on lilliputian hallucinations. Jan Dirk Blom. Neuroscience & Biobehavioral Reviews, March 4 2021. https://doi.org/10.1016/j.neubiorev.2021.03.002

Rolf Degen's take: https://twitter.com/DegenRolf/status/1367705194562265090

Highlights

• Lilliputian hallucinations are not as harmless as traditionally assumed.

• Their etiology is diverse, with CNS pathology accounting for a third of the cases.

• Therefore, in most cases auxiliary investigations are advisable.

• Treatment is directed at the underlying cause.

• A failure of size constancy may explain part of the underlying mechanism.

Abstract: Lilliputian hallucinations concern hallucinated human, animal or fantasy entities of minute size. Having been famously described by the French psychiatrist Raoul Leroy in 1909, who wrote from personal experience, to date they are mentioned almost routinely in textbooks of psychiatry, albeit with little in-depth knowledge. I therefore systematically reviewed 145 case reports and case series comprising 226 case descriptions, concluding that lilliputian hallucinations are visual (61%) or multimodal (39%) in nature. In 97% of the cases, they are perceived as grounded in the actual environment, thus indicating involvement of higher-level regions of the perceptual network subserving the fusion of sensory and hallucinatory content. Perceptual release and deafferentiation are the most likely underlying mechanisms. Etiology is extremely diverse, with schizophrenia spectrum disorder, alcohol use disorder and loss of vision accounting for 50% of the cases and neurological disease for 36%. Recovery was obtained in 62% of the cases, whereas 18% of the cases ended in chronicity and 8% in death. Recommendations are made for clinical practice and future research.

Keywords: Alcohol hallucinosisCharles Bonnet syndromeentity experienceintoxicationmultimodal hallucinationpsychedelicssize constancy

4.6. Pathophysiology

MRI scans of patients experiencing lilliputian hallucinations indicate the involvement of primary and secondary visual cortex in their mediation (Chen and Liu, 2011; Vacchiano et al., 2019). Since systematic localizing studies are lacking and other MRI studies have, moreover, shown involvement of frontal, parietal and mesencephalic regions (Walterfang et al., 2012; Hirakawa et al., 2016), it is hard to tell which parts of the visual network are of primary importance. Nonetheless, a special role would seem to be reserved for visual association cortex and - especially in compound hallucinations - higher-level cortical association networks. This follows from the complexity and remarkable embeddedness of the hallucinations in the observer’s surroundings. Individual phenomenological characteristics also hint at a role for lower levels of the visual network, even down to V4 in color perception and V5 in motion perception. Although much remains to be elucidated here, a case can also be made for the applicability of two different pathophysiological mechanisms, i.e. deafferentiation and perceptual release. It should be noted, though, that neither of these mechanisms are likely to produce lilliputian hallucinations without the involvement of higher-level perceptual networks.

4.6.1. Deafferentiation

The deafferentiation model states that loss of peripheral sensory input can lead to spurious activity of central networks normally involved in processing that input. It is thus applicable to closed-eye hallucinations, hypnagogic hallucinations and Charles Bonnet syndrome (ffytche et al., 1998), including cases of hemi- and quadrantanopsia. Moreover, this mechanism has been described in Parkinson’s disease, with the lilliputian hallucinations appearing at dusk (i.e., crepuscular hallucinations) (Montassut & Sauguet, 1943). A special case, and an exception in fact, is one involving a patient with blindness of the right eye and hemianopia of the left eye, who experienced lilliputian hallucinations in the one remaining hemifield, seeing ‘little men’ populating the world that he could still actually see (Camus, 1911), which is thus exemplary of perceptual release.

4.6.2. Perceptual release

The perceptual release model states that endogenously mediated perceptual material, which during our waking hours normally remains below the threshold of consciousness, can break through the surface and be projected into the outside world. Also referred to as ‘dream intrusion’, the model characterizes lilliputian hallucinations as a matter of ‘dreaming while awake’. A crucial difference with dreaming though, is that in this case only part of the perceived environment is replaced by hallucinatory content. The model is not only applicable to hypnopompic hallucinations (Trénel, 1926), but also to the majority of other cases falling outside the ‘sensory deprivation’ category (Table 1). Here, the term peduncular hallucinosis refers to a concept developed by Lhermitte and van Bogaert regarding the alleged involvement of the midbrain and pons (Kosty et al., 2019). However, whether these mesencephalic regions are indeed responsible for (all instances of) perceptual release is as yet uncertain. What is more, the phenomenological characteristics of peduncular hallucinosis far transcend those of lilliputian hallucinations.

4.6.3. Adjuvant models

Four other models likely play minor roles in the mediation of lilliputian hallucinations. The first is the peripheral model, in which peripheral hallucinations are attributed to intraocular pathology such as glaucoma, cataract and entoptic phenomena. Glaucoma and cataract are thought to act via the intermediate process of visual loss and hence deafferentiation, whereas entoptic phenomena (e.g., floaters, vitreous opacities) serve as points de repères for the development of lilliputian hallucinations, much in the way of visual pareidolia. With eyes closed, simple visual phenomena such as photopsia and Eigengrau are thought to fulfill similar catalytic functions (Ahlenstiel, 1954). All such entoptic phenomena result in hallucinations of the swiveling projective type, i.e. those moving along with the observer’s gaze, which were reported in 2% of the cases only. The second adjuvant model is the reperception model, which holds that hallucinations may borrow their content from memory. In the literature reviewed, I found a rare example concerning a 54-year-old woman who saw faithful miniature versions of people she had known at school, who were moreover saying things that made perfect sense in the light of her memories (Jacome, 1999). Then there is the irritation model, which states that epileptic activity may selectively stimulate brain regions responsible for hallucinatory activity. As we saw, epilepsy was diagnosed in only 7% of the cases. Moreover, the complex (and sometimes compound) nature of lilliputian hallucinations suggests that even in epilepsy, these phenomena are probably mediated via the intermediate process of perceptual release. Finally, the psychodynamic model needs mention. It suggests that lilliputian hallucinations are typically experienced by people with playful, imaginative or regressive personalities. In exceptional cases, the model may perhaps help to explain the appreciation or even the content of hallucinations, but I ask myself how the observations might contribute to our understanding of their mediation.

4.7. Size constancy

An intriguing and as yet little understood aspect of lilliputian hallucinations is the minute size of the beings featuring in them. In 1922 a crucial role was already suggested for failed size constancy in their mediation (Salomon, 1922), where size constancy is the capacity to assess the size of objects as uniform at different ranges. Emmert’s law states that this ability depends on depth cues and on the size of the retinal image (Millard et al., 2020). It was thus proposed that, in lilliputian hallucinations, this mechanism goes awry when hallucinated material is ‘projected’ on surfaces in confined spaces, making the ensuing image appear unnaturally small (Salomon, 1922). Even if correct, though, that explanation would only be applicable to hallucinations of the cinematic projective type, which make up a mere 1% of all cases. Only recently, lesion studies and electrophysiological studies have started to shed light on the neural correlates of size constancy. Drawing on two cases of brain infarction, it has been suggested that size constancy is a function of (the forceps major of) the dominant side of the brain (Hong et al., 2010). In an animal study with macaques, tungsten microelectrodes have been inserted into single neurons located in V4, leading to the identification of object-size coding neurons, i.e. neurons that compute object size independently of retinal size (Tanaka & Fujita, 2015). Similar experiments with cats yielded evidence for the involvement of specialized neurons in V1 (Pigarev & Levichkina, 2016), while a study in humans with steady-state visually evoked potentials found even more robust results for the latter region (Chen et al., 2019). While the exact mechanisms subserving size constancy remain to be elucidated, evidence is thus accumulating that specific neurons in early visual cortex play a role in the process, perhaps even without having to rely on retinal size, as was indicated by one study (Tanaka & Fujita, 2015), which is obviously also the case in lilliputian hallucinations.

4.8. Limitations

The number of published case descriptions of lilliputian hallucinations is small and most stem from historical sources. Even though I summarized the original case descriptions as faithfully as possible, it was not always clear whether they had been complete in the first place, and whether neurological patients were not also alcoholics, for instance, or vice versa. As a consequence, there is room for doubt about the accuracy of several of the conclusions drawn in this review. Moreover, since reports on lilliputian hallucinations were sometimes buried in the literature on Charles Bonnet syndrome, where they often serve as examples of the syndrome’s characteristic complex visual hallucinations, I may have missed an unknown number of case descriptions for my systematic review.