Moyamoya disease (MMD) is a potentially life-threatening disease characterized by stenosis or occlusion of intracranial arteries. The exact cause of MMD is still unknown, but it is believed to be at least partly hereditary with mainly maternal transmission and a high female-to-male incidence ratio. The distal segments of the internal carotid arteries, along with the proximal middle and anterior cerebral arteries, are most frequently affected (1). While patients may present with unilateral stenosis, bilateral involvement can develop over time. In rarer instances, the posterior circulation may also be affected, particularly the posterior cerebral artery (2). Pathological MMD findings include gradual narrowing of the blood vessels, in the absence of major inflammation or atherosclerosis (3, 4). This progressive stenosis allows the development of compensatory networks of thin collateral vessels, after which the disease takes its name. On angiographic images, these compensatory networks resemble to “puffs of smoke”, hence the term “Moyamoya”, which in Japanese means hazy like a puff of smoke in the air. When this condition is idiopathic it is referred to as MMD. Conversely, when the vascular findings are observed alongside an associated condition, e.g. neurofibromatosis type 1, it is termed Moyamoya syndrome (MMS) (5). Both MMD and MMS share similar symptoms including transient ischemic attacks, seizures, headaches, cognitive impairments, and can result in the occurrence of ischemic or hemorrhagic stroke in both pediatric and adult populations (6). However, MMD typically presents with bilateral vascular involvement (although there may be a predominant side), whereas MMS may exhibit unilateral or bilateral vessel involvement (7). Treatment options primarily aim to restore blood flow through surgical interventions. Although MMD is less common than MMS and is considered a rare condition, its prevalence has been reported to be as high as 3.1 per 100,000 individuals in Japan, where it is most commonly observed (8, 9). MMD exhibits a bimodal age presentation, with the first peak typically occurring during the first decade of life and the second during the fourth decade (10).

Four main MMD types have been identified based on their clinical presentations: the hemorrhagic, the epileptic, the transient ischemic attack (TIA), and the ischemic infarct types (8). The hemorrhagic type is more prevalent in the fourth decade of life, and is related to intracranial hemorrhage, arising from the rupture of the delicate network of collateral vessels, primarily manifesting as intraparenchymal bleeding, although it can occasionally result in intraventricular or subarachnoid bleeding (11). The epileptic type is characterized by seizures, which have been found to be the initial manifestation in up to 34% of all cases in a prospective study of 23 MMD patients (12). The most common MMD type is the ischemic one. The ischemic infarct type accounts for approximately 95% of younger patients (first decade of life), who often suffer also from the TIA type. Especially in these two last types, it can be hypothesized that certain triggers, such as crying, coughing, blowing, hyperventilation, or stress can lead to ischemic symptoms due to reduced carbon dioxide levels and vasoconstriction (8, 13). A decrease in cerebral blood supply, and any organic cerebral insult in general, can contribute to the development of psychiatric symptoms. The localization of cerebral lesions may also be specifically linked with the type of associated psychiatric symptoms. Indeed, recent research has placed a growing emphasis on understanding and characterizing the psychiatric manifestations of MMD (14). However, there is a notable gap in evidence regarding the association between demographic characteristics, such as gender, patterns of MMD (e.g., hemispheric laterality of MMD involvement or the presence of stroke), and distinct psychiatric symptoms or diagnostic presentations. Recognizing these characteristics and associations holds the potential to improve the diagnosis of MMD in psychiatric patients, enhance the precision of psychiatric symptom screening in MMD patients, and optimize psychiatric care for individuals with MMD. Indeed, a significant proportion of patients with MMD may have psychiatric comorbidity (15), and could benefit the most from appropriate interventions that account for the peculiar etiopathogenesis of their psychiatric symptoms.

This pre-registered systematic review and meta-analysis provides the first systematic assessment and characterization of the existing literature on psychiatric manifestations in MMD patients, with the primary aim of improving clinical detection and characterization of psychiatric presentations of this rare pathology. By suggesting some potential pathophysiological hypotheses concerning the origin of psychiatric symptoms in MMD, and by discussing the diagnostic, therapeutic, and prognostic implications tied to the recognition and characterization of specific patterns of psychiatric symptoms in MMD, the review outlines a roadmap for future research and clinical directions, paving the way for the development of targeted psychiatric treatments and interventions for MMD patients.

1588 non-duplicate studies were identified through database and manual searching and they were screened, as described in Figure 1. The final sample included 41 articles reporting psychiatric symptoms in MMD patients. Of these, 29 were case reports, and 12 were other kind of observational studies. No study was excluded due to missing information. All the extracted data is detailed in Supplementary Tables 2, 3, and in the following paragraphs.

This pre-registered systematic review and meta-analysis investigating, for the first-time, psychiatric comorbidity of MMD yielded 41 relevant studies, of which 29 were case reports. The quality of the included studies was generally fair with a low risk of bias.

Notwithstanding some limitations detailed in Section 4.1, both quantitative and qualitative syntheses of our systematic review and meta-analysis highlight that psychiatric symptoms are highly prevalent in MMD patients, and that relatively often these symptoms are the only presenting symptoms leading to an MMD diagnosis. In such cases, none of the patients described in the included case reports had a prior psychiatric history. These findings are in line with two previous narrative reviews (14, 15) and emphasize the vital significance of improving recognition and care of psychiatric symptoms in individuals with MMD, as their diagnosis and management can be particularly challenging. To strive towards this goal, we cautiously propose some preliminary hypotheses on the characteristics of psychiatric symptoms in MMD, clinical implications, and potential psychotropic treatments that have shown promise in managing psychiatric symptoms in MMD patients, based on the findings of our review.

The review did not highlight gender differences in psychiatric symptoms in MMD patients. However, our included sample demonstrated a marginal preponderance of females, in agreement with the well-documented higher prevalence of MMD in women (56). Concerning age differences, qualitative synthesis suggests that anxio-depressive symptoms are more common in adults, while, in children and adolescents, psychosis, ADHD symptoms, emotional lability, and impaired concentration with poor academic performance have been described.

The psychiatric symptom most frequently described in the case reports was psychosis, while in other types of studies anxio-depressive symptoms were the most common ones. This may reflect a reporting bias among case reports, as well as the fact that psychotic manifestations are usually more striking and temporally better defined than anxio-depressive ones and may thus be linked more easily with MMD. Notwithstanding these considerations, intriguing patterns emerge from the review. We will discuss them in the following section, although further, mechanistic, research is needed on these topics, as mentioned in the Limitations (Section 4.2.).

First of all, the fact that psychotic symptoms were more common in MMD patients with predominant involvement of the left hemisphere (i.e. the dominant one, considering that none of the included articles reported left-handed individuals), suggests a potential association between the location of MMD pathology and specific psychiatric manifestations. The left hemisphere is prominently involved in language processing, logical reasoning, and semantic memory. Thus, damage to this region could contribute to cognitive distortions and impaired reality testing, which are often observed in individuals with psychosis, delusions, or schizophrenia (57). For instance, we may speculate that lesions affecting Broca’s area or the left temporoparietal junction may disrupt the seamless integration of language and social cues, leading to misinterpretations and distorted beliefs characteristic of psychotic thinking (58). Furthermore, the left hemisphere plays a pivotal role in top-down cognitive control and attentional processes. Lesions or reduced blood flow in this region may compromise an individual’s ability to filter out irrelevant stimuli and maintain focused attention. As an hypothesis, this may potentially lead to aberrant saliency assignment, which is one of the proposed physiopathological mechanisms for psychosis since increased susceptibility to external stimuli being misinterpreted or distorted fuels the development of delusions or other psychotic phenomena (59, 60). In fact, a recent fMRI study highlighted aberrant functional connectivity between the saliency network and the left, but not the right, hippocampus in bipolar disorder, which may represent a marker of vulnerability to psychosis (61). Similarly, Liu and colleagues (62) highlighted structural and functional alterations of the left thalamus in schizophrenic patients, and these impairments correlated with symptoms. Thalamus involvement in psychosis is, indeed, well-established and supported by fMRI studies (63). Finally, a systematic review on neural underpinnings of delusional symptoms in schizophrenia and Alzheimer’s disease patients reported lateralized findings of gray matter reductions in the left claustrum (64), reinforcing the notion of left hemisphere involvement in the genesis of psychotic symptoms.

While these observations align with the hypothesis of left hemisphere involvement in psychotic symptoms in some populations, a systematic review in patients with post-stroke psychosis found that right hemisphere lesions were more common (in around 40% of the cases, against 11% with a different pattern of lesions). However, no quantitative synthesis was performed in this systematic review, and about half of the included case reports did not give information on lesions localization, limiting the reliability of these results (65). Considering our finding of an increased association of psychosis with left hemisphere lesions in MMD patients, it may be hypothesized that psychosis associated with MMD has different characteristics and mechanisms compared to poststroke psychosis, even in the presence of brain lesions.

This hypothesis is rooted in the proposed physiopathology of the generation of psychiatric symptoms, and specifically of psychotic outbursts in MMD. Indeed, as mentioned in the Introduction, it has been proposed that these are most often acute or acutely exacerbated, commonly in response to certain triggers, such as crying, coughing, blowing, hyperventilation, or stress. These can lead to reduced carbon dioxide levels and vasoconstriction, which is more deleterious in the brain regions already significantly affected by MMD (8, 13). A consequent transitory decrease in cerebral blood supply, especially in such vulnerable regions, may contribute to the development of panic (34, 66) or even psychotic symptoms, which have been consistently associated with cerebral hypoperfusion (67). On the other hand, in post-stroke psychosis, the affected brain regions are permanently impaired, and the symptoms may result from more stable reorganization of brain networks or neurotransmitters imbalance. This hypothetical mechanism gains further credence from the observation that electroencephalograms in MMD patients, particularly among pediatric patients, frequently reveal persistent build-up and rebuilding phenomena. This phenomenon entails the continued presence of high-amplitude slow waves beyond 30 seconds following hyperventilation. Notably, these EEG changes often normalize post-surgery and show a strong correlation with cerebral ischemic events and vascular reserve, a link substantiated by various imaging techniques such as cerebral angiography or single-photon emission-computed tomography (10, 34, 68). Indeed, the context of the occurrence of psychotic symptoms in some of the included case reports support the existence of a link between certain triggers that may lead to cerebral vasoconstrictions and psychotic symptoms. For instance, Klasen et al. (29) suggest a causal link between MMD and psychotic symptoms in a 12-year-old boy, given the improbability of these two events co-occurring independently at his age, without any previous family or personal psychiatric history, the specific involvement of the left hemisphere and left basal ganglia, and the physical exertion accompanied by hyperventilation preceding the psychosis outburst. This is in agreement with another case report describing a woman who presented recurrent exacerbation of psychotic symptoms and hallucinations after physical exercise for a decade before being diagnosed with MMD, and before finally responding to clozapine (43). Of note, her pattern of cerebral hypoperfusion was clearly predominant in the left hemisphere (43). Concerning hallucinations in particular, another case report (14) argues that they were linked to hypoperfusion in the territory of the posterior cerebral arteries. This hypothesis aligns with the observation that psychosis appears to be a more frequent manifestation in children and young adults with MMD (15), who more often present with the ischemic/TIA type of MMD, potentially exposing them to a higher likelihood of psychotic outbreaks in association to transitory exacerbations of cerebral hypoperfusion. However, MMD patients may develop psychiatric symptoms as a result of MMD at any point of their illness. In fact, insight from longitudinal studies underscores the evolving susceptibility of brain regions to hemodynamic insufficiency as MMD progresses, with a shift observed from inner to outer cortical structures and from anterior to posterior territories. This dynamic evolution potentially contributes to the development of new psychiatric manifestations over time, even in otherwise stable MMD (69).

Undoubtedly, MMD exposes patients to a myriad of other factors that may increase their risk of developing psychiatric symptoms. While our focus revolves around symptom characterization and mechanistic details are beyond the scope of this review, it is noteworthy to mention, that higher levels of stress and traumatic experiences may significantly elevate the risk of anxio-depressive disorders. The independent effect of stroke should also be considered. Indeed, stroke has been associated with different psychiatric disorders, such as depression, potentially due to serotoninergic disturbances (70), and, rarely, psychosis (65).

Frequent cerebral ischemic and hemorrhagic lesions in MMD may partially contribute to the augmented vulnerability of these patients to psychiatric symptoms. However, almost half of MMD patients with psychiatric symptoms described in the included case reports did not exhibit stroke nor TIA, aligning with previous literature (15). Additionally, our analysis demonstrates that MMD patients who experienced strokes did not exhibit a different propensity for psychosis compared to those without strokes.

Finally, we showed that MMD patients exposed to stroke were less likely of being diagnosed with MMD following isolated psychiatric symptoms only, and that the presence of neurological symptoms made it more likely for the MMD diagnosis to precede psychiatric symptoms. This pattern is logical, as MMD is often more easily diagnosed following strokes or neurological symptoms, which is not the case for instances where MMD presents solely with isolated psychiatric symptoms (15).

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

LS: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. CM: Writing – review & editing, Data curation. AW: Conceptualization, Supervision, Writing – review & editing. MS: Conceptualization, Data curation, Supervision, Writing – review & editing.