Introduction

Migraine is a common disorder, often characterised by episodes of pulsating unilateral headaches, as defined in the International Classification of Headache Disorders, third edition (ICD-3; Headache Classification Committee of the International Headache Society (IHS) [1]). The Global Burden of Disease study ranked migraine as the sixth-most prevalent and among the top 10 most disabling diseases, affecting approximately 14.4% of individuals worldwide [2]. Migraine can be diagnosed as with or without aura, which refers to completely reversible central nervous system disturbances [1]. Six types of aura symptoms are described, namely, involving visual, sensory, speech and/or language, motor, brainstem or retinal disturbances. Most common are visual aura symptoms, examples of which can include zigzag figures at a central fixation point, scotoma involving blind spots or scintillating scotoma characterised by flickering lights or colours.

While complex visual phenomena are currently also included as aura symptoms, hallucinations or other unusual sensory experiences in non-visual modalities have not been accounted for. “Sensory aura” implies that other senses may be involved, but only somatic events involving numbness or tinging, or pins-and-needles have been well described; auditory, olfactory, gustatory, or complex tactile experiences are seemingly absent within existing diagnostic criteria. Yet literature has suggested that these latter non-visual experiences are in fact, not uncommon in migraine [3,4,5]. As the ICHD-3 nosology has focused on visual and somatic aura symptoms like those outlined, most empirical studies have been in these areas [6]. In contrast, research regarding the phenomenology of complex hallucinations or other unusual perceptual experiences involving non-visual modalities has been relatively limited.

Understanding these experiences has clear clinical implications for awareness, diagnosis and management for both individuals with migraine and their clinicians. Nosologically, it is important that the disease classification captures typical migraine symptoms as well as less common aura experiences, to ensure accurate and inclusive diagnosis. Having a comprehensive understanding of the range of symptoms experienced in migraine is also important for investigations of complex neurobiological mechanisms contributing to onset. Finally, precision medicine has been highlighted as a critical future step in migraine diagnosis and intervention [7], for which a complete picture of symptoms at hand is needed.

The present study aimed to bridge these gaps in existing literature, by systematically reviewing and synthesising past research regarding multisensory hallucinations and other unusual sensory experiences in migraine. The following primary research questions were posed: (i) what was the phenomenology of complex multisensory hallucinations as potential migraine aura symptoms? And (ii) what other multisensory unusual perceptual experiences, distinct from complex hallucinations, may be identified? A third aim was to explore whether there were notable patterns or trends in migraine symptoms, including frequency, as associated with manifestation of these hallucinatory or other unusual sensory events, within our included samples.

Methods

The review adhered to guidelines described in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [8]. The study protocol was pre-registered on the International Prospective Register Reviews (PROSPERO; CRD42022298603).

Search strategy and selection criteria

A systematic search was conducted across PsycINFO and Web of Science, for peer-reviewed publications in English, from 1980 to the present, with a cut-off date of 1st February 2023. Search terms were focused on two primary keywords/phrases: “hallucinations” or “unusual sensory experiences”, based on non-visual domains in the context of migraine. Optimised permutations of search terms were used (see Table A in Supplementary materials for search syntax by database).

Articles were screened based on the following inclusion criteria: (i) main outcome variable(s) involving hallucinations or unusual sensory experiences outside of established aura symptoms or known prodromal or postdromal indicators (e.g. photophobia or phonophobia). Exclusion criteria were as follows: (i) no diagnosis of migraine; (ii) participants aged under 12 years old; (iii) significant neurological comorbidity causing migraine (e.g. tumours or brain injury) and (iv) replicated datasets (for which the single most appropriate paper was included to avoid duplication).

Study selection and data extraction

After removing duplicates, YL screened the title, abstract and keywords of retrieved publications, and excluded irrelevant studies. A review of full-text articles was then conducted independently by CY and YL using the inclusion/exclusion criteria outlined. Discussions with senior author WLT were conducted to resolve any discrepancies in study selection between YL and CY. The search strategy yielded a total of 1745 records, from which n = 48 (case studies = 19, group design studies = 29) met inclusion criteria in the final stage. Figure 1 shows the PRISMA flow diagram.

Fig. 1
figure 1

PRISMA flow diagram

Full size image

Data extraction was jointly performed by CY and YL and comprised study identifiers (author(s), year of publication), participant characteristics (e.g. age, sex, subgroup sizes), phenomenological descriptions of established aura symptoms (including visual and sensory aura), complex hallucinatory experiences, prodromal or postdromal symptoms and other unusual sensory experiences (otherwise unaccounted for).

Risk of bias assessment

A risk of bias assessment was conducted by YL using the JBI Critical Appraisals Checklist for Case Studies (JBI-CAC; Munn [9]) and Newcastle–Ottawa Quality Assessment Scale (NOQAS; [10] for group design studies, adapted for our purpose (see Tables C and D in Supplementary materials). One item was added to the JBI-CAC to assess descriptive information, as this review was focused primarily on the phenomenology of hallucinations or unusual sensory experiences in the context of migraine; items relevant to interventions or treatment were removed. Two items were added to the NOQAS to assess the validity of hallucination and migraine symptom assessments. Within each of the seven domains, 0–2 points were awarded, based on how well each response satisfied the criterion under assessment (nb. only 0–1 points could be awarded for the item “Representativeness of the cases” for NOQAS). The unweighted summed scores provide an indication of the overall quality and bias of each study. Categorical ratings were devised as follows: Excellent 10–12/12, good 7–9/12, fair 4–6/12 and poor ≤ 3/12 for the JBI-CAC and NOQAS.

Results

Results from case and group design studies were extracted and summarised separately, owing to disparate aims and variables employed across these distinct types of empirical research.

Multisensory hallucinatory and other unusual sensory experiences: case studies

Descriptive phenomenology of multisensory and other unusual sensory experiences in migraine extracted from case studies (n = 19) is shown in Table 1. To contextualise these findings, we opted to present participant characteristics, comprising age, sex, neurological and psychiatric comorbidities as well as associated migraine symptoms and their timelines, where available. A wide participant age range was observed (12–72 years old), with slightly more males (56%) than females overall. Neurological and psychiatric comorbidities varied, ranging from focal lesions to epilepsy for the former and bipolar or depressive disorders for the latter. Yet significant neurological or psychiatric histories (including psychotic illness) were excluded for about half of the included studies (n = 10).

Table 1 Multisensory hallucinatory and other unusual sensory experiences in migraine (n = 19 case studies)
Full size table

Established aura symptoms were reported in all but five studies; visual symptoms usually involved some form of photopsia, scotoma and/or fortification spectra, whereas motor symptoms commonly related to paresis, numbness or tingling, typically in the perioral region or limbs, which could transpire uni- or bilaterally. The nature of complex hallucinatory experiences was diverse, and could include events in one or more of the major sensory modalities, namely, hearing voices or other sounds, smelling pleasant or unpleasant odours, accompanied by unusual tastes at times, or feeling touch sensations across the skin (as well as visual hallucinations), and more obscure experiences, such as Lilliputian hallucinations or perceived time travel. Documented phenomenological characteristics (e.g. frequency, level of conviction, functional impairment) were relatively scarce, though distress was explicitly noted in relation to auditory and (some) visual hallucinations. If present, prodromal or postdromal sensory symptoms commonly involved photophobia and/or phonophobia (n = 12). Other unusual sensory experiences, comprising osmophobia or hypoesthesia among others, were also described (n = 4).

In terms of associated migraine symptoms, age of onset or length of illness, headache frequency and duration, as well as pain quality, localisation and intensity all exhibited significant variation across participants. Hallucinatory (or unusual sensory) experiences mostly took place prior to headache onset, though some participants reported having these experiences during and/or after their headache (possibly depending on the nature of the sensory event).

Multisensory hallucinatory and other unusual sensory experiences: group design studies

Details of established aura symptoms and the phenomenology of complex hallucinations and other unusual sensory experiences in migraine extracted from group design studies (n = 29) are presented in Table 2. Some unusual sensory experiences (such as ear fullness) outlined in Table 2 have been noted as established symptoms for other headache disorders, but are currently omitted from the diagnostic criteria for migraine [1]. Neurological and psychiatric comorbidities, associated migraine symptoms and relationships between migraine symptoms and aura/hallucinations have been placed in Table B of Supplementary materials, due to limited group design studies presenting these data.

Table 2 Multisensory hallucinatory and other unusual sensory experience in migraine (n = 29 empirical studies)
Full size table

Of the group design studies, n = 17 reported established aura symptoms, including visual and sensory aura and vertigo. Descriptions of complex hallucinatory experiences varied, and commonly included auditory distortions (n = 3), tactile hallucinations (n = 3), Alice in Wonderland syndrome (n = 2), olfactory hallucinations (n = 2) or “changes” (n = 2). Auditory distortions included the inability to hear one’s own voice [11], or perceiving “uncharacteristic” or “distorted” noises [12]. Tactile hallucinations could involve feelings such as red, hot ears bilaterally [13], or one’s head being hooped by a ribbon, pressed by a stone, or feeling as though one was wearing a hat [6]. Coenesthetic hallucinations lasting the duration of the headache were also reported, where participants felt as though their brain was coming out of their head [11]. Reports of Alice in Wonderland syndrome included altered or distorted perception, involving one’s body size, weight or position in space [14]. Other distortions included extrapersonal misperceptions, such as out-of-body experiences and derealisation, typically lasting the duration of the headache [11]. Macrosomatognosia (e.g. feeling “like [they are] growing so big [their] head would push through the roof”) and microsomatognosia (e.g. feeling as though “[they are] shrinking and becoming so small [they] would disappear behind the steering wheel”) were reported, in addition to aschematia and time distortion [11]. Olfactory hallucinations could include the smell of foods/drinks (e.g. vanilla, melon, coffee, banana, bitter almonds), burning (e.g. burnt wood, smoke), chemicals/metals (e.g. gas, metallic, sulphur) or other negative smells (e.g. rotten meat). Individuals usually experienced the same olfactory hallucinations, with roughly half having sudden onset and remission. Most of these experiences lasted 3–10 min and were considered “unlikable” [4].

The overall frequency of migraine was found to be positively associated with the number of different types of unusual sensory symptoms experienced [14]. Compared to those without aura, individuals who experienced aura symptoms often reported a higher proportion of hallucinations or unusual sensory experiences, such as olfactory [15] or tactile [6] hallucinations, bodily temperature changes [15], “distorted” auditory, olfactory, gustatory or tactile events, or sensory hypersensitivities, such as osmophobia (n = 14; [5, 16]. Jürgens et al. [14] also reported that those with established aura had higher frequencies of autokinesis, corona phenomenon, cinematographic vision, metamorphopsia, dyschromatopsia, illusionary spread and synaesthesia. However, individuals without migraine aura still reported tactile [6] or olfactory hallucinations [15], as well as other unusual sensory experiences [5], including sensory hypersensitivities [15, 16], albeit at lower levels. People with migraine with osmophobia judged odours as less pleasant than those without olfactory hypersensitivity and reported significantly more painful and frequent headaches, as well as increased numbers of odour-induced migraines [17]. Some individuals with migraine also had more difficultly detecting and discriminating odours relative to healthy controls [18]. Aural fullness was outlined by limited studies (n = 3). The most common postdromal and prodromal symptoms were photophobia (n = 19) and phonophobia (n = 16). A large proportion of people with migraine with aura experienced headaches after the onset of the aura, of which most individuals experienced headache during the aura, or within 30 min after the cessation of the aura [19].

Overall, complex hallucinations and unusual sensory experiences in the context of migraine were diverse and numerous. Established visual aura were described as more fascinating than worrisome [11], but there were limited data regarding distress and functional impairment for the other sensory modalities. Distortions or hypersensitivities across diverse visual, auditory, olfactory, tactile and gustatory senses were evident, suggesting the presence of significant and varied sensory experiences, beyond typical established aura symptoms.

Discussion

The current study represented a systematic review summarising the phenomenology of multisensory and other unusual perceptual experiences in migraine, outside the criteria of well-established aura, especially involving non-visual domains. Data were extracted from eligible case and group design studies, evaluated for quality based on requisite risk of bias assessments.

In response to our first research question, complex hallucinations were vividly described in more than half of all included studies (n = 29), involving auditory, olfactory, tactile, coenesthetic, gustatory modalities as well as other peripheral senses. Experiences within these dominant sensory modalities typically implicated hearing voices (or other sounds), smelling or tasting varied odours or flavours, or burning or touch sensations on one’s skin. Other unusual sensory experiences comprised Lilliputian hallucinations or Alice in Wonderland syndrome involving distortion of bodily perception, or less well-described perceptions implicating aural fullness, time distortions, out-of-body encounters or other anomalous events. However, existing accounts largely lacked systematic phenomenological information encompassing physical (e.g. frequency, duration), cognitive (e.g. controllability, insight) and emotional (e.g. distress, functional interference) experiential facets. This has rendered it difficult to draw consistent conclusions about the characteristics of these multisensory events. Other than isolated anecdotal accounts of distress associated with some forms of auditory (or visual) hallucinations, it was also hard to decipher which modalities or aspects were significantly associated with negative emotional outcomes or interfered with daily functioning in a way that may benefit from therapeutic intervention. One reason for this lack of attention could be that such aura symptoms are typically considered transient or reversible phenomena (though this does not preclude their potential for evoking distress, or persisting for a significant period). In fact, some migraine experiences could be mistaken for severe intoxication, psychiatric disorders or even a medical emergency. A possible exception was the study of olfactory hallucinations in migraine, which was methodically managed across a large-scale study providing rich phenomenological detail [4]. This study demonstrated that most participants did consider their olfactory experiences as unpleasant (although typical episodes were brief with sudden onset and comprising recurrent content).

In response to our second research question, other unusual sensory experiences (distinct from complex hallucinations) were noted in also more than half of all included studies (n = 29). Known prodromal or postdromal symptoms were reported, but there were also descriptions involving osmophobia (n = 15), as well as relatively rarer symptoms, which have yet to be classified as being nosologically associated with migraine. Focused investigations are therefore needed to determine how these symptoms should best be accounted for within a migraine disease framework. To address our exploratory research question, a cursory analysis of participant characteristics revealed no systematic patterns in terms of the age, sex or neurological and psychiatric comorbidities of affected individuals, and group design studies were also unable to establish consistent associations between hallucinatory experiences and specific migraine symptoms. Finally, our risk of bias assessments demonstrated that the limited conclusions drawn from our review were supported by the inclusion of studies largely of high quality (n = 40). Nevertheless, the presence of a number of lower quality studies (n = 11), coupled with limited data points pertaining to our variables of interest, signifies that future replication remains imperative.

Clinical implications and recommendations

Despite the limited conclusive findings, there are still useful clinical lessons to be drawn from our review. First, multisensory complex hallucinations and/or unusual perceptual experiences, outside of the visual realm, affecting people with migraine are not uncommon. This can take place before, during and/or after the headache episode (if present), the temporal order of which could be specific to the individual or dependent on the type of experience. The unexpected onset and nature of these events can bring about surprise or even bewilderment in some affected persons. Furthermore, involvement of specific modalities (e.g. auditory) and/or content (e.g. unpleasant) is likely to trigger significant distress in some patients, although there could be other experiential facets (e.g. physical, cognitive or emotional) also directly related to negative emotional outcomes (but these remain unknown). By bringing attention to and directly asking about these experiences, it is hoped that improving clinician–patient awareness and communication may convey some benefit. In some cases, there may be scope for therapeutic intervention to manage patient distress. This can be in the form of basic psychoeducation, normalising to patients that such sensory perceptions can be part of the migraine episode, with no underlying sinister implications, and not to feel overly alarmed, where possible. Or in instances where the distress is overwhelming, appropriate referrals to mental health professionals for psychological treatment may be necessary.

Second, there are nosological implications worth further consideration. There is some incongruency in that complex visual hallucinations are currently captured within the diagnostic category of migraine with HIS [1], but complex hallucinations in other non-visual modalities are not accounted for anywhere in the current classification system. Visual hallucinations (and other visual symptoms) fall within the class of visual aura, but although there exists the label of sensory aura, this mostly pertains to sensations of numbness or pins-and-needles, neglecting complex non-visual hallucinations, even though these have clearly been reported by individuals with migraine. A similar case can be made for other unusual perceptual experiences, especially involving osmophobia, or even aural fullness. Akin to photophobia and phonophobia, these sensory hypersensitivities seemingly precede or ensue headache episodes as prodromal or postdromal symptoms, respectively, but have yet to be nosologically recognised within existing diagnostic criteria. Related to this, it is worth noting that the present study variously referred to aura, prodromal/postdromal and interictal symptoms, without fully separating these out at times. While it is important to delineate aura symptoms from those experienced during the prodrome/postdrome and interictal phases, the temporal course of these events is often complex and difficult to assess, with significant overlaps noted. Therefore, future research should seek to better delineate which symptoms fall under each (or a combination) of these phases for clearer nosological distinction (also see “Directions for future research” section).

The current findings may contribute to improved definitions of the neural effects of migraine and may aid in establishing migraine as more than a headache condition. Following these preliminary suggestions from the current review, added investigations to corroborate a robust empirical foundation supporting these recommendations, alongside open scientific debate by prevailing experts, would be constructive. Should an affirmative consensus be achieved, implementing these changes in the next classification revision would ensure nosological clarity and consistency moving forward. However, it remains important that any defined aura symptom should still meet ICHD-3 criteria.

Study limitations

Our systematic review was subject to several limitations. Owing to the nature of our research questions, we had excluded studies focusing solely on sensory perceptual experiences already established within existing diagnostic criteria for migraine [1]. This largely comprised not only visual symptoms, whether in the form of complex hallucinations or other anomalous perceptions (e.g. photopsia, scotoma or fortification spectra), but could also include sensorimotor symptoms (e.g. paresis, numbness/tingling) as well as known prodromal indicators (e.g. photophobia or phonophobia). Such investigations did not contribute towards our consolidation of current knowledge to address existing gaps in the study of atypical non-visual sensory experiences in migraine (nb. the study of visual symptoms is evidently noteworthy, with n = 15 already identified in our search, suggesting it would necessitate a separate review). That being said, for completeness, we did extract sensory information pertaining to the visual realm, but only when this was presented alongside, in studies focusing on non-visual symptoms. Thus, any conclusions drawn from the current review need to be interpreted within this lens.

We also elected not to include an analysis of medication effects in our review, as this information had not been systematically collected across many of the included studies, meaning that any conclusions would likely have been biased in some way. However, given the possible impact of pharmacological drugs on unusual sensory experiences, this issue is deserving of detailed scrutiny under more opportune circumstances. Akin to this, a lack of consistent examination/associations between hallucinatory and migraine symptoms meant that we were unable to decipher meaningful patterns or arrive at more definitive conclusions in this regard (also see Table B in Supplementary materials). Phenomenological information, including frequency of modality-specific hallucinations as well as linking of timelines of unusual perceptual events relative to headache onset, especially key for mechanistic understandings, was only reported sporadically. We were thus unable to organise the extracted data in more meaningful ways, for instance, by ordering via magnitude of increasing frequency, according to prodromal versus postdromal occurrence, or clustering in line with similarity of experiences. On the other hand, our extensive search strategy, coupled with inclusion of a mix of study types, aimed at shedding light on this hitherto neglected area in migraine, is a key strength. Some of the most vivid descriptions of non-visual complex hallucinations tended to stem from case reports, whereas group design studies facilitated meaningful comparisons across subgroups with relatively robust participant numbers.

Directions for future research

There are some fruitful avenues worth pursuing in terms of future research endeavours. Evidently, there needs to be more comprehensive investigations into all modes of unusual sensory experiences, including complex hallucinations, in the context of migraine. The prevailing emphasis on visual symptoms (although possibly more prevalent) at the expense of neglecting other sensory modalities denotes a bias that needs to be overcome via a more inclusive approach. This would comprise an examination of related phenomenological parameters (i.e. physical, cognitive and emotional) as well as patient characteristics and associated migraine symptoms. Given its less established nature in migraine studies, the validity of phenomenological inquiry of multisensory hallucinations may be buttressed by borrowing from appropriate existing measures in the psychosis field, or if necessary, creating new tools not premised on the dominance of visual sensory symptoms. Based on such studies, we may begin to elucidate which modalities/types of experiences are most prevalent and distressing, which subgroups of people with migraine are most likely to be affected (according to specific demographic or clinical features) and which key migraine symptoms can serve as prognostic indicators and/or therapeutic targets.

Related to this, tracking the temporal interaction of unusual sensory and migraine symptoms could be especially revealing. This could involve not just the timing of sensory and migraine symptoms relative to one another but also how such temporal ordering may evolve over time within individuals, potentially as a sign of migraine disease progression. From an etiological perspective, the involvement of multiple sensory processing systems before, during and/or after a migraine episode likely implicates the contribution of diverse neurobiological regions [20, 21]. Studying its overt manifestation in the form of unusual sensory perception may lead to improved mechanistic understandings that could culminate in therapeutic breakthroughs, especially if definitive neurobiological structures or causes can be isolated. Ensuing knowledge from these proposed lines of research will likely in turn bring about direct nosological and treatment benefits. Finally, taking into account the anomalous nature of such perceptual events, consideration should be given to investigating potential stigma in terms of reporting these phenomena as well as coping strategies. This could help raise awareness of these lesser-known migraine symptoms, possibly mistakenly associated with mental illness.

Conclusions

The current systematic review aimed to provide a comprehensive documentation of multisensory complex hallucinations and other unusual perceptual experiences (beyond the visual domain) in the context of migraine. It was concluded that these experiences affect a certain proportion of people with migraine and vary widely in their presentation. This can range from typical everyday perceptions (e.g. smelling food scents) to extraordinary and vivid sensory events across one or more senses (e.g. hearing/seeing human interactions), or even border on the bizarre. It is likely that certain experiential modalities and/or facets are directly linked to patient distress, though added research is necessary to corroborate which phenomenological characteristics are most liable. This research contributes to a greater understanding of the neural effects of migraine and may help establish it as more than a headache condition. From a clinical perspective, increased clinician–patient awareness is vital to manage distress (where necessary), and scientifically, related nosological issues warrant further consideration. An inclusive management of these unusual sensory symptoms may be key to offering a holistic therapeutic approach in migraine. Given that precision medicine has been identified as a key next step in migraine diagnosis and management treatment [7], recognition of these understudied multisensory experiences is warranted.