Definition of PE

Within this review, PE is used to describe only hallucinations and delusions which occur at a sub-clinical level to those in the general population. This term is commonly used, however it creates categorical division of a continuum. The concept of the ‘extended psychosis-phenotype’, hypothesises that all psychotic phenomena exist on a continuum (van Os, Reference van Os2003; van Os, Linscott, Myin-Germeys, Delespaul, & Krabbendam, Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009). At one end are individuals with psychotic disorders (severe and frequent hallucinations/delusions). At the other are individuals in the general population who experience occasional mild hallucinations/delusions. Evidence supports this fully dimensional scale of psychosis, showing evidence of shared environmental, genetic and neurobiological factors (DeRosse & Karlsgodt, Reference DeRosse and Karlsgodt2015; Johns & Van Os, Reference Johns and Van Os2001; van Os et al., Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009).

Despite this robust evidence, research on subclinical psychotic phenomena commonly use categorical definitions e.g. PE or those at clinical high risk for psychosis (CHR). Both are forms of subclinical psychotic symptoms, and share common symptomology (Fusar-Poli et al., Reference Fusar-Poli, Borgwardt, Bechdolf, Addington, Riecher-Rössler, Schultze-Lutter and Yung2013; Kelleher & Cannon, Reference Kelleher and Cannon2011). Proposed delineations exist, such as those at CHR showing help-seeking behaviour (van Os & Linscott, Reference van Os and Linscott2012). However, CHR have been found in non-help-seeking populations (Staines et al., Reference Staines, Gajwani, Gross, Gumley, Lawrie, Schwannauer and Uhlhaas2021; Uhlhaas et al., Reference Uhlhaas, Gajwani, Gross, Gumley, Lawrie and Schwannauer2017), and those reporting PE show help-seeking behaviour (Armando et al., Reference Armando, Nelson, Yung, Saba, Monducci, Dario and Girardi2012), making this differentiation insufficient. Similarly, distress about symptoms as a delineation (Johns et al., Reference Johns, Kompus, Connell, Humpston, Lincoln, Longden and Larøi2014), is vulnerable to individual differences (Coughlan et al., Reference Coughlan, Healy, Humphries, Clarke, Kelleher and Cannon2020).

Additionally, PE as defined within this review does not discuss subclinical negative symptoms, cognitive deficits, thought disorder or affective disorder as PE. This is as in the literature, they are not commonly examined as ‘PE’. When both are included they are often examined as a wider phenomenon e.g. ‘psychotic experiences and negative symptoms (PENS)’ in (Pain et al., Reference Pain, Dudbridge, Cardno, Freeman, Lu, Lundstrom and Ronald2018).

Occurrence of psychotic experiences in the general population

The projected lifetime risk of PE is 7.8% (McGrath et al., Reference McGrath, Saha, Al-Hamzawi, Alonso, Andrade, Borges and Kessler2016). However, prevalence differs across a lifetime, at around 5–7% in adulthood (Linscott & van Os, Reference Linscott and van Os2013; van Os et al., Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009; Yates et al., Reference Yates, Lång, Peters, Wigman, McNicholas, Cannon and Kelleher2021), but a significantly higher prevalence in children (17%) and moderately higher in adolescence (8%) (Kelleher et al., Reference Kelleher, Connor, Clarke, Devlin, Harley and Cannon2012). A study of PE across a lifetime observed this declining trend, finding the lowest prevalence in later adulthood (⩾ 70 years) of 3% (Yates et al., Reference Yates, Lång, Peters, Wigman, McNicholas, Cannon and Kelleher2021). However (Linscott & van Os, Reference Linscott and van Os2013) examined methodological variables and observed reporting type (self-report/clinical interview) accounted for 19.7% of the variance in prevalence rates of PE. Similarly, different types of PE may have differing prevalence rates, e.g. a systematic review of auditory PE found there were equal prevalence in children (12.7%) and adolescence (12.4%) (Maijer, Begemann, Palmen, Leucht, & Sommer, Reference Maijer, Begemann, Palmen, Leucht and Sommer2018).

Less is known about incidence, one review yielded estimates between 2.5% and 3% median incidence rate across a few (k = 6) studies (Linscott & van Os, Reference Linscott and van Os2013; van Os et al., Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009). Patterns for difference in age are consistent in adulthood; one large-scale adult sample showed a decline in incidence from early to late adulthood (Tien, Reference Tien1991). Intriguingly, one study in a child and adolescent sample found the incidence was higher in adolescence than early childhood (Sullivan et al., Reference Sullivan, Kounali, Cannon, David, Fletcher, Holmans and Zammit2020). A systematic review of hallucinatory PE found a similar result of incidence rate per-person year range of 0.1–1.3% in children and adolescents (Rubio, Sanjuán, Flórez-Salamanca, & Cuesta, Reference Rubio, Sanjuán, Flórez-Salamanca and Cuesta2012). Reasons for differences between prevalence and incidence are currently unknown, although methodological differences must be considered (Linscott & van Os, Reference Linscott and van Os2013; Maijer et al., Reference Maijer, Hayward, Fernyhough, Calkins, Debbané, Jardri and Bartels-Velthuis2019a).

For most individuals (~80%) PE are transient, (Linscott & van Os, Reference Linscott and van Os2013; McGrath et al., Reference McGrath, Saha, Al-Hamzawi, Alonso, Bromet, Bruffaerts and Kessler2015). For the remaining (~20%) PE may recur. Individuals who experience multi-modal PE are more likely to experience recurring PE (McGrath et al., Reference McGrath, Saha, Al-Hamzawi, Alonso, Bromet, Bruffaerts and Kessler2015). Studies commonly describe PE which occur repeatedly as ‘persistent’ (e.g. van Os et al., Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009). However, even in those with repeated incidence of PE, these symptoms remit for significant lengths of time (McGrath et al., Reference McGrath, Saha, Al-Hamzawi, Alonso, Bromet, Bruffaerts and Kessler2015). Therefore this paper has opted for the term ‘recurring PE’ which we believe better reflects the phenomena.

Challenges in researching psychotic experiences

Two key challenges of PE research: phenomenology & measurement.

Hallucinations and delusions vary widely in phenomenology, e.g. auditory hallucination and paranoid ideation are both common PE (Armando et al., Reference Armando, Nelson, Yung, Ross, Birchwood, Girardi and Fiori Nastro2010; Coughlan et al., Reference Coughlan, Healy, Humphries, Clarke, Kelleher and Cannon2020). Additionally, subtypes of PE may represent different trajectories, with evidence finding bizarre experiences and perceptual abnormalities were associated with higher distress, depression and worse functioning (Armando et al., Reference Armando, Nelson, Yung, Ross, Birchwood, Girardi and Fiori Nastro2010). Some PE subtypes are more associated with suicidal behaviour (Hielscher et al., Reference Hielscher, DeVylder, Hasking, Connell, Martin and Scott2020, Reference Hielscher, DeVylder, Hasking, Connell, Martin and Scott2021).

There is significant heterogeneity in measurement tools; (Lee et al., Reference Lee, Chan, Chang, Lee, Hui and Chen2016) reported that across 76 studies, 41 different assessment tools were used. There is some suggestion that self-report measures may generate more false-positive ratings of PE (Schultze-Lutter et al., Reference Schultze-Lutter, Renner, Paruch, Julkowski, Klosterkötter and Ruhrmann2014; van Os et al., Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009), and have been found to report higher rates of PE than clinical interview (Healy et al., Reference Healy, Brannigan, Dooley, Coughlan, Clarke, Kelleher and Cannon2019a). One study comparing the Development-and-Well-Being Assessment (DAWBA) self-report tool to a clinical interview observed that the sensitivity (73.8%), specificity (77.1%), and the negative predictive value (96.3%) were good, but positive predictive value (26.8%) was low (Gundersen et al., Reference Gundersen, Goodman, Clemmensen, Rimvall, Munkholm, Rask and Jeppesen2019). Additionally, sensitivity (10.0–43.1) of symptom subtype was poor for the DAWBA (Gundersen et al., Reference Gundersen, Goodman, Clemmensen, Rimvall, Munkholm, Rask and Jeppesen2019). However, self-report and clinically validated PE show similar outcomes (Healy et al., Reference Healy, Brannigan, Dooley, Coughlan, Clarke, Kelleher and Cannon2019a), and ‘false positive ratings’ of PE were still found to be associated with psychiatric outcomes (van der Steen et al., Reference van der Steen, Myin-Germeys, van Nierop, ten Have, de Graaf, van Dorsselaer and van Winkel2019).

Additionally, self-report measures without information on duration, frequency, or conviction risk misidentifying events. One large-scale sample found that of those identified with self-report PE, 11% also met criteria for CHR, and a subsample (n = 15) met criteria for overt psychosis (Moriyama et al., Reference Moriyama, van Os, Gadelha, Pan, Salum, Manfro and Drukker2019). Questions on auditory hallucinations appear to be the most valid single-item measure of PE within child and adolescent populations (Kelleher & Cannon, Reference Kelleher and Cannon2011). Although (Leiderman, Reference Leiderman2011) found visual hallucinations were more frequent in their sample, and the DAWBA found visual hallucinations had a higher predictive value (Gundersen et al., Reference Gundersen, Goodman, Clemmensen, Rimvall, Munkholm, Rask and Jeppesen2019). Overreliance on single-items has limitations; it restricts research to a symptom subtype, and a single item endorsement is likely to produce false negatives.

Heritability and genetics

Sex differences have been examined to mixed results, one systematic review did observe higher rates in males (van Os et al., Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009), although a second review and a large international sample both failed to replicate this (Linscott & van Os, Reference Linscott and van Os2013; McGrath et al., Reference McGrath, Saha, Al-Hamzawi, Alonso, Bromet, Bruffaerts and Kessler2015). One systematic review for heritability & genetics and PE exists, but was limited by a small number of studies (k = 13) (Ronald & Pain, Reference Ronald and Pain2018). Single-nucleotide polymorphism (SNP)-studies failed to find evidence of heritability for PE (Ronald & Pain, Reference Ronald and Pain2018). Genome-wide association studies (GWAS) (Ortega-Alonso et al., Reference Ortega-Alonso, Ekelund, Sarin, Miettunen, Veijola, Järvelin and Hennah2017; Pain et al., Reference Pain, Dudbridge, Cardno, Freeman, Lu, Lundstrom and Ronald2018) found evidence of genome-wide significant association, but neither was able to independently replicate these findings (Ronald & Pain, Reference Ronald and Pain2018). Newer studies on GWAS have expanded these findings showing genetic overlap between PE and psychiatric disorders including schizophrenia and bipolar disorder (Barkhuizen, Pain, Dudbridge, & Ronald, Reference Barkhuizen, Pain, Dudbridge and Ronald2020; Legge et al., Reference Legge, Jones, Kendall, Pardiñas, Menzies, Bracher-Smith and Walters2019). Poly-genetic risk scores were the most studied in the systematic review (k = 10) (Ronald & Pain, Reference Ronald and Pain2018) which showed moderate evidence for overlapping genetic risk for schizophrenia and PE. Investigations into gene-environment interactions indicate some SNP and risk haplotypes interact with environmental factors of bullying (Cristóbal-Narváez et al., Reference Cristóbal-Narváez, Sheinbaum, Rosa, Ballespí, de Castro-Catala, Peña and Barrantes-Vidal2016), drug use (Zammit, Owen, Evans, Heron, & Lewis, Reference Zammit, Owen, Evans, Heron and Lewis2011), and childhood trauma (Ramsay et al., Reference Ramsay, Kelleher, Flannery, Clarke, Lynch, Harley and Cannon2013).

Pre/perinatal complications

Higher birth weight has been associated with a reduced risk of PE (Thomas et al., Reference Thomas, Harrison, Zammit, Lewis, Horwood, Heron and Gunnell2009), and lower birthweight with increased risk (Drakesmith et al., Reference Drakesmith, Dutt, Fonville, Zammit, Reichenberg, Evans and David2016). Maternal infections, negative life events, cannabis use and smoking during pregnancy have been identified as a risk factor for PE (Dorrington et al., Reference Dorrington, Zammit, Asher, Evans, Heron and Lewis2014; Fine et al., Reference Fine, Moreau, Karcher, Agrawal, Rogers, Barch and Bogdan2019; Zammit et al., Reference Zammit, Thomas, Thompson, Horwood, Menezes, Gunnell and Harrison2009). Maternal binge drinking has not been linked with PE (Gregersen, Dreier, & Strandberg-Larsen, Reference Gregersen, Dreier and Strandberg-Larsen2020), but large weekly consumption of alcohol was (Zammit et al., Reference Zammit, Thomas, Thompson, Horwood, Menezes, Gunnell and Harrison2009). Both studies used large samples and collected drinking data during pregnancy, allowing for solid comparability. Older paternal age is associated with risk for PE (Foutz & Mezuk, Reference Foutz and Mezuk2014). Winter season of birth shows a moderate risk for PE in meta-analysis (Córdova-Palomera et al., Reference Córdova-Palomera, Calati, Arias, Ibáñez, Moya, Ortet and Fañanás2015).

Early development and health factors

Children exposed to Epstein-Barr Virus or atopic conditions in childhood increase risk of adolescent PE (Khandaker, Pearson, Zammit, Lewis, & Jones, Reference Khandaker, Pearson, Zammit, Lewis and Jones2014). Similarly, childhood dysregulation of inflammatory markers (interleukin 6 and C-reactive protein) and certain complement proteins (C1RL, C5, CFH) have been identified as an early biomarker for later development of PE (Föcking et al., Reference Föcking, Sabherwal, Cates, Scaife, Dicker, Hryniewiecka and Cotter2019; Khandaker et al., Reference Khandaker, Pearson, Zammit, Lewis and Jones2014; Mongan et al., Reference Mongan, Föcking, Healy, Susai, Heurich and Wynne2021). These results suggest a role for the early-life and immune system development in vulnerability to PE.

Cognitive dysfunction

Ross, McKay, Coltheart, and Langdon (Reference Ross, McKay, Coltheart and Langdon2015) in a systematic review found that individuals with subclinical delusions reported worse data gathering abilities in ‘jumping to conclusions’ (JTC) measures. A much broader systematic review and meta-analysis conducted by Livet, Navarri, Potvin, and Conrod (Reference Livet, Navarri, Potvin and Conrod2020) found PE in the general population were strongly associated with aberrant salience, moderate trends in external attribution biases and attention to threat, and weak evidence for JTC and belief inflexibility. However heterogeneity between studies was significant (Livet et al., Reference Livet, Navarri, Potvin and Conrod2020). Other studies have observed additional cognitive dysfunction such as motor speed, and IQ (Carey et al., Reference Carey, Dooley, Gillan, Healy, Coughlan, Clarke and Cannon2019; Horwood et al., Reference Horwood, Salvi, Thomas, Duffy, Gunnell, Hollis and Harrison2008), and evidence suggests these can be observed from early infancy (Carey et al., Reference Carey, Healy, Perry, Gillan, Whitehouse, Cannon and Lin2021). However given heterogeneity between studies further analysis is needed, and a greater understanding of why these results differ between groups.

Neuroanatomical development

A range of subtle neuro-anatomical differences have been observed in those reporting PE. These include, global and localised differences in grey matter volume and white matter integrity (Dooley et al., Reference Dooley, O'Hanlon, Healy, Adair, McCandless, Coppinger and Cannon2020; Satterthwaite et al., Reference Satterthwaite, Vandekar, Wolf, Bassett, Ruparel, Shehzad and Gur2015), and widespread structural and functional differences across the frontal, temporal, & parietal lobes, and basal ganglia, all regions strongly implicated in psychosis (Drakesmith et al., Reference Drakesmith, Dutt, Fonville, Zammit, Reichenberg, Evans and David2016; Jacobson et al., Reference Jacobson, Kelleher, Harley, Murtagh, Clarke, Blanchard and Cannon2010; Modinos, Ormel, & Aleman, Reference Modinos, Ormel and Aleman2010; Okada et al., Reference Okada, Yahata, Koshiyama, Morita, Sawada, Kanata and Kasai2018). Additionally, widespread functional dysconnectivity (default mode dysfunction in particular, but with increasing interest in the motor network) and network efficiency deficits (Drakesmith et al., Reference Drakesmith, Caeyenberghs, Dutt, Zammit, Evans, Reichenberg and Jones2015; Jacobson McEwen et al., Reference Jacobson McEwen, Connolly, Kelly, Kelleher, O'Hanlon, Clarke and Garavan2014; Karcher, O'Brien, Kandala, & Barch, Reference Karcher, O'Brien, Kandala and Barch2019; Orr, Turner, & Mittal, Reference Orr, Turner and Mittal2014; O'Neill et al., Reference O'Neill, Carey, Dooley, Healy, Coughlan, Kelly and Cannon2020). However, it is currently unclear whether the differences in brain structure and functioning precede or follow the onset of PE.

Urbanicity

Individuals living in densely populated areas have been shown to have a higher risk of PE (Linscott & van Os, Reference Linscott and van Os2013; van Os et al., Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009). Research on air pollution found it was a significant modifier and accounted for 60% of the variance (Newbury et al., Reference Newbury, Arseneault, Beevers, Kitwiroon, Roberts, Pariante and Fisher2019). Additionally, urbanicity and PE in low and middle-income countries failed to show an association (DeVylder et al., Reference DeVylder, Kelleher, Lalane, Oh, Link and Koyanagi2018b). One large-scale study in China found rural upbringing was associated with elevated risk for PE (Wang et al., Reference Wang, Wang, Li, Zhang, Wei, Deng and Li2019). This study identified social stress and economic deprivation was higher in rural China, both of which show increased risk of PE (DeVylder et al., Reference DeVylder, Koyanagi, Unick, Oh, Nam and Stickley2016). Negative perception of neighbourhood has also been linked to PE (Newbury et al., Reference Newbury, Arseneault, Caspi, Moffitt, Odgers, Baldwin and Fisher2017). These findings suggest culturally dependent differences in urban/rural life, pollution and perception, may underpin the mechanisms by which urbanicity can elevate risk of PE.

Ethnic minority status

Research on ethnic minority status (EMS) & migrant status and PE indicate that there are differences based on EMS, with high discrepancies between groups. One meta-analysis found certain EMS groups showed higher PE in certain regions (Leaune et al., Reference Leaune, Dealberto, Luck, Grot, Zeroug-Vial, Poulet and Brunelin2019), even when sociodemographic variables were accounted for. However the high heterogeneity between different EMS groups suggested discrimination and deprivation of certain EMS groups might better explain this finding (Leaune et al., Reference Leaune, Dealberto, Luck, Grot, Zeroug-Vial, Poulet and Brunelin2019). A second meta-analysis supported this by showing that accounting for psychological stress and clinical diagnosis, in addition to sociodemographic variables, attenuated these findings (Tortelli et al., Reference Tortelli, Nakamura, Suprani, Schürhoff, Van der Waerden, Szöke and Pignon2018).

A systematic review of discrimination observed experiencing discrimination increased risk for PE, particularly paranoia (Pearce, Rafiq, Simpson, & Varese, Reference Pearce, Rafiq, Simpson and Varese2019). A second US-based review observed negative police interactions increased risk for PE and other mental disorders in black Americans (McLeod, Heller, Manze, & Echeverria, Reference McLeod, Heller, Manze and Echeverria2020), although only one study specifically examined PE (DeVylder et al., Reference DeVylder, Jun, Fedina, Coleman, Anglin, Cogburn and Barth2018a). Both (Leaune et al., Reference Leaune, Dealberto, Luck, Grot, Zeroug-Vial, Poulet and Brunelin2019; Tortelli et al., Reference Tortelli, Nakamura, Suprani, Schürhoff, Van der Waerden, Szöke and Pignon2018) failed to find evidence that migrant status alone indicated increased risk of PE. One meta-analysis found residential concentration of own ethnic identity was shown to be a protective factor against PE among EMS (Bécares, Dewey, & Das-Munshi, Reference Bécares, Dewey and Das-Munshi2018).

Socio-economic status

Lower socio-economic status (SES) is associated with higher rates of PE (Loch et al., Reference Loch, Chianca, Alves, Freitas, Hortêncio, Andrade and Rössler2017; Mamah, Mutiso, & Ndetei, Reference Mamah, Mutiso and Ndetei2021; van Os et al., Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009). As with urbanicity, research has begun to examine the mechanisms which might produce these differences; One study found social disadvantage, discrimination, and unstable housing situations increased SES predictive value (Veling & Adriaanse, Reference Veling and Adriaanse2013). Another examining PE in children found neighbourhood deprivation and poverty were both significantly associated with PE (Karcher, Schiffman, & Barch, Reference Karcher, Schiffman and Barch2021b).

Trauma

Trauma such as sexual, physical, or emotional abuse increases risk of PE (Daalman et al., Reference Daalman, Diederen, Derks, van Lutterveld, Kahn and Sommer2012; Fisher et al., Reference Fisher, Schreier, Zammit, Maughan, Munafò, Lewis and Wolke2013). Studies have demonstrated a dose-response rate between traumatic experiences and PE (Coughlan & Cannon, Reference Coughlan and Cannon2017; Shevlin et al., Reference Shevlin, Murphy, Read, Mallett, Adamson and Houston2011). A recent meta-analysis observed distinctive pathways between trauma and subsequent PE subtype, driven by differing psychological processes (e.g. post-traumatic symptomology, attachment and social cognition) (Bloomfield et al., Reference Bloomfield, Chang, Woodl, Lyons, Cheng, Bauer-Staeb and Lewis2021). However this analysis was limited by the significant number (k = 14 of 22) of cross-sectional designs of included studies.

There is substantial evidence of a link between bullying and PE (Fisher et al., Reference Fisher, Schreier, Zammit, Maughan, Munafò, Lewis and Wolke2013; Wolke, Lereya, Fisher, Lewis, & Zammit, Reference Wolke, Lereya, Fisher, Lewis and Zammit2014). Within a longitudinal study of adolescents, Kelleher et al. (Reference Kelleher, Keeley, Corcoran, Ramsay, Wasserman, Carli and Cannon2013) observed a bidirectional relationship between bullying and rates of PE. Evidence indicates multi-victimisation results in higher rates of PE (Arseneault et al., Reference Arseneault, Cannon, Fisher, Polanczyk, Moffitt and Caspi2011). Research into paranoid delusions has shown direct causal links to negative affect, bullying, peer difficulties, and negative behaviours on social media (Bird, Evans, Waite, Loe, & Freeman, Reference Bird, Evans, Waite, Loe and Freeman2019).

Psychopathology

A large systematic review of 15 studies found a majority showed internalising and externalising behaviours in childhood and adolescence increase risk of PE (Gin, Stewart, & Jolley, Reference Gin, Stewart and Jolley2021). Research using a large longitudinal sample of children and adolescents has demonstrated a bi-directional relationship between psychopathology and PE (Healy, Coughlan, Clarke, Kelleher, & Cannon, Reference Healy, Coughlan, Clarke, Kelleher and Cannon2020). Insomnia and excessive daytime somnolence have been linked to higher rates of PE (Barton, Kyle, Varese, Jones, & Haddock, Reference Barton, Kyle, Varese, Jones and Haddock2018; Reeve, Sheaves, & Freeman, Reference Reeve, Sheaves and Freeman2015). Evidence suggests a bi-directional relationship between insomnia and PE (Reeve, Nickless, Sheaves, & Freeman, Reference Reeve, Nickless, Sheaves and Freeman2018).

Smoking tobacco or cannabis

Tobacco use has been shown to increase risk for PE (Bhavsar et al., Reference Bhavsar, Jauhar, Murray, Hotopf, Hatch, McNeill and MacCabe2018a). Shakoor et al. (Reference Shakoor, Zavos, McGuire, Cardno, Freeman and Ronald2015) found use of cannabis and the presence of PE had shared environmental risk factors. A systematic review found younger age, and frequent cannabis use are associated with higher reports of PE (Ragazzi, Shuhama, Menezes, & Del-Ben, Reference Ragazzi, Shuhama, Menezes and Del-Ben2018). Maternal cannabis use during late-term pregnancy increased psychosis-proneness (Fine et al., Reference Fine, Moreau, Karcher, Agrawal, Rogers, Barch and Bogdan2019). Kuepper et al. (Reference Kuepper, van Os, Lieb, Wittchen, Höfler and Henquet2011) demonstrated continued cannabis use was associated with a greater risk of recurring PE.

Risk factors for recurring psychotic experiences

Although evidence supports that recurring PE are associated with poorer outcomes (Bhavsar et al., Reference Bhavsar, Dorrington, Morgan, Hatch, McGuire, Fusar-Poli and Hotopf2021; Rimvall et al., Reference Rimvall, Wolf, Olsen, Skovgaard, Clemmensen, Oxholm and Jeppesen2021), currently less is known on recurring risk factors. Evidence has suggested that recurring PE are not associated with family history of psychosis (Karcher et al., Reference Karcher, Loewy, Savill, Avenevoli, Huber, Makowski and Barch2021a; Thapar et al., Reference Thapar, Heron, Jones, Owen, Lewis and Zammit2012), although one study did observe a moderate trend (Karcher et al., Reference Karcher, Loewy, Savill, Avenevoli, Huber, Makowski and Barch2021a). Studies examining sex have found more women report recurring PE; using latent growth modelling techniques higher numbers of female participants were in their ‘persisting’ & ‘increasing’ PE groups (Mackie, Castellanos-Ryan, & Conrod, Reference Mackie, Castellanos-Ryan and Conrod2011; Thapar et al., Reference Thapar, Heron, Jones, Owen, Lewis and Zammit2012), and other forms of analysis have found a similar gender disparity (DeVylder, Lehmann, & Chen, Reference DeVylder, Lehmann and Chen2015; Fonville et al., Reference Fonville, Drakesmith, Zammit, Lewis, Jones and David2019). Early developmental problems including birth complications, maternal smoking and infections were associated with higher rates of recurring PE (Rössler et al., Reference Rössler, Riecher-Rössler, Angst, Murray, Gamma, Eich and Gross2007; Thapar et al., Reference Thapar, Heron, Jones, Owen, Lewis and Zammit2012). Cognitive deficits are underexplored in the literature, but currently show no association with recurring PE (Kalman, Bresnahan, Schulze, & Susser, Reference Kalman, Bresnahan, Schulze and Susser2019).

EMS was associated with recurring PE, one study in an American sample (Calkins et al., Reference Calkins, Moore, Satterthwaite, Wolf, Turetsky, Roalf and Gur2017) found higher rates of recurring PE in African American participants (76.6% v. 42.5% in controls). However a second study failed to find these differences (DeVylder et al., Reference DeVylder, Lehmann and Chen2015). Other environmental factors including urbanicity and SES were not found to be related (Bourque, Afzali, O'Leary-Barrett, & Conrod, Reference Bourque, Afzali, O'Leary-Barrett and Conrod2017; Peters et al., Reference Peters, Ward, Jackson, Morgan, Charalambides, McGuire and Garety2016; Wigman et al., Reference Wigman, Van Winkel, Raaijmakers, Ormel, Verhulst, Reijneveld and Vollebergh2011).

Behavioural measures such as frequent cannabis use, stressful-life events, trauma, victimisation and bullying have been associated with recurring PE (Bourque et al., Reference Bourque, Afzali, O'Leary-Barrett and Conrod2017; Mackie et al., Reference Mackie, O'Leary-Barrett, Al-Khudhairy, Castellanos-Ryan, Struve, Topper and Conrod2013; Sun et al., Reference Sun, Xue, Zhang, Guo, Hu, Li and Rosenheck2017; Wigman et al., Reference Wigman, Van Winkel, Raaijmakers, Ormel, Verhulst, Reijneveld and Vollebergh2011). Cougnard et al. (Reference Cougnard, Marcelis, Myin-Germeys, Graaf, Vollebergh, Krabbendam and Os2007) further reported that recurring PE were associated with a higher number of cumulative environmental risk factors. Markers of mental ill health including more severe symptomology, higher medication prescriptions, insomnia, higher reported rates of anxiety and depression, and low functioning have all been linked to recurring PE (Calkins et al., Reference Calkins, Moore, Satterthwaite, Wolf, Turetsky, Roalf and Gur2017; Reeve et al., Reference Reeve, Sheaves and Freeman2015; Yamasaki et al., Reference Yamasaki, Usami, Sasaki, Koike, Ando, Kitagawa and Sasaki2018).

Despite a number of known risk factors, in a recent scoping review, Kalman et al. (Reference Kalman, Bresnahan, Schulze and Susser2019) failed to find any consistent evidence of risk factors for recurring PE. Modelling techniques to develop predictive models of recurring v. transient PE found a similar limitation using current known risk factors (AUC = 0.66, 7.4% of variance, Sensitivity = 68.8%; Specificity = 54.4%) (Steenkamp et al., Reference Steenkamp, Tiemeier, Blanken, Hillegers, Kushner and Bolhuis2021), although the model did differentiate recurring PE from control.

Psychotic disorders

A systematic review of PE found adolescent PE had a 4-fold increased risk for developing psychotic disorder (Healy et al., Reference Healy, Brannigan, Dooley, Coughlan, Clarke, Kelleher and Cannon2019a), and this effect remained even after controlling for childhood psychopathology (Fisher et al., Reference Fisher, Schreier, Zammit, Maughan, Munafò, Lewis and Wolke2013). Recurring PE examined in a longitudinal study design (8.4 years) found recurring PE had increased risk of psychotic disorders relative to transient PE (Dominguez, Wichers, Lieb, Wittchen, & van Os, Reference Dominguez, Wichers, Lieb, Wittchen and van Os2011).

Non-psychotic disorders

A meta-analysis of longitudinal prospective studies robustly demonstrated the relationship between non-psychotic disorders and PE (Kaymaz et al., Reference Kaymaz, Drukker, Lieb, Wittchen, Werbeloff, Weiser and van Os2012). Healy et al. (Reference Healy, Brannigan, Dooley, Coughlan, Clarke, Kelleher and Cannon2019a, Reference Healy, Coughlan, Williams, Clarke, Kelleher and Cannon2019b) found childhood PE was associated with a 3-fold increased risk for mental disorder. Conversely, Varghese et al. (Reference Varghese, Scott, Welham, Bor, Najman, O'Callaghan and McGrath2011) found the presence of major depressive or anxiety disorder was associated with 4-fold and 3-fold elevated risk for subsequent PE. One systematic review found PE subtype showed different associations, observing paranoia PE, but not grandiosity, was highly associated with anxiety (Bird, Waite, & Freeman, Reference Bird, Waite and Freeman2018). These associations are further elevated for those with recurring PE (Chan et al., Reference Chan, Lee, Chan, Pang, Wong, Hui and Chen2021; Downs, Cullen, Barragan, & Laurens, Reference Downs, Cullen, Barragan and Laurens2013). Co-occurring PE and mental disorders is well documented (Downs et al., Reference Downs, Cullen, Barragan and Laurens2013; Kelleher et al., Reference Kelleher, Connor, Clarke, Devlin, Harley and Cannon2012; Laurens, Downs, Cullen, Barragan, & To, Reference Laurens, Downs, Cullen, Barragan and To2012). Evidence shows a majority of adolescents with PE meet criteria for at least one lifetime non-psychotic mental disorder, and had particularly elevated risk of psychiatric multi-morbidity (Armando et al., Reference Armando, Nelson, Yung, Ross, Birchwood, Girardi and Fiori Nastro2010; Laurens et al., Reference Laurens, Downs, Cullen, Barragan and To2012). The presence of PE in those with mental disorders were associated with more severe symptoms of anxiety and depression, and showed slower recovery rates (Knight et al., Reference Knight, Russo, Stochl, Croudace, Fowler, Grey and Perez2020).

Psychotic experiences are associated with subsequent psychopathology, with the presence of PE being associated with higher rates of substance disorders, internalising & externalising symptoms in adolescence (Cederlöf et al., Reference Cederlöf, Kuja-Halkola, Larsson, Sjölander, Östberg, Lundström and Lichtenstein2017), and early adulthood (Carey et al., Reference Carey, Gillan, Healy, Dooley, Campbell, McGrane and Cannon2021). PE are associated with escalating trajectory of psychopathology (Iorfino et al., Reference Iorfino, Scott, Carpenter, Cross, Hermens, Killedar and Hickie2019). Recurring PE is associated with worse psychopathology (Downs et al., Reference Downs, Cullen, Barragan and Laurens2013).

Healthcare needs

Research suggests PE are associated with an elevated risk for future psychiatric diagnosis, mental health service use, pharmacological treatment, healthcare costs and health-related quality of life (Rimvall et al., Reference Rimvall, van Os, Verhulst, Wolf, Larsen, Clemmensen and Jeppesen2020, Reference Rimvall, Wolf, Olsen, Skovgaard, Clemmensen, Oxholm and Jeppesen2021). These findings were shown using the Copenhagen Child Cohort 2000, a longitudinal study design that at baseline represented 1 in 10 children born in Denmark (Olsen et al., Reference Olsen, Rask, Elberling, Jeppesen, Clemmensen, Munkholm and Skovgaard2020). This study had the advantage of access to The Danish National Health Service Register, allowing for comprehensive analysis of healthcare costs and use. In particular, the presence of PE and a co-occurring mental disorder were associated with a greater risk for these measures (Rimvall et al., Reference Rimvall, van Os, Verhulst, Wolf, Larsen, Clemmensen and Jeppesen2020). Similar results have been found in other independent samples (Bhavsar et al., Reference Bhavsar, Dorrington, Morgan, Hatch, McGuire, Fusar-Poli and Hotopf2021), with one (Karcher et al., Reference Karcher, Loewy, Savill, Avenevoli, Huber, Makowski and Barch2021a) observing that persistence and distress related to PE both increased mental health service use. A moderately sized (k = 13) meta-analysis observed individuals with PE showed a 2-fold increase for mental health service use compared to healthy controls (Bhavsar et al., Reference Bhavsar, McGuire, MacCabe, Oliver and Fusar-Poli2018b). Additionally, studies have observed the presence of PE is associated with more medical conditions (e.g. asthma, arthris, hearing/visual problems) (Moreno et al., Reference Moreno, Nuevo, Chatterji, Verdes, Arango and Ayuso-Mateos2013). Most concerning, a longitudinal study in a large sample (n = 15 049) found that the presence of PE was associated with a 5 year median shorter lifespan (Sharifi et al., Reference Sharifi, Eaton, Wu, Roth, Burchett and Mojtabai2015), with neoplasms and death by suicide being significantly higher than those without PE at baseline.

Suicidal behaviour

Psychotic experiences are strongly associated with suicidal ideation, attempts and death by suicide (Yates et al., Reference Yates, Lång, Cederlöf, Boland, Taylor, Cannon and Kelleher2019). Systematic reviews and meta-analyses have indicated that those with PE were at a 2–3 fold increased risk of suicidal behaviours (Honings, Drukker, Groen, & van Os, Reference Honings, Drukker, Groen and van Os2016a; Yates et al., Reference Yates, Lång, Cederlöf, Boland, Taylor, Cannon and Kelleher2019). Recurring PE are associated with an even greater risk of suicide behaviour relative to transient-PE (Connell et al., Reference Connell, Betts, McGrath, Alati, Najman, Clavarino and Scott2016). One study of adolescents found depression mediated the relationship between PE and suicide (Nunez et al., Reference Nunez, Campos, Spencer, Faúndez, Fresno and Bravo2020), and a review expanded on this observing significant mediating roles of mental disorders, psychological distress and environmental factors (Hielscher et al., Reference Hielscher, DeVylder, Hasking, Connell, Martin and Scott2021).

Mediators of psychotic experiences

To date substantially less is known on the mediators of PE. Research has identified a number of potential mediators between PE and risk/outcome factors. These mediators include parent-child conflict, self-concept, internalising problems, coping strategies and social and community cohesion (Crush, Arseneault, Jaffee, Danese, & Fisher, Reference Crush, Arseneault, Jaffee, Danese and Fisher2018; Dhondt, Staines, Healy, & Cannon, Reference Dhondt, Staines, Healy and Cannon2022; Healy et al., Reference Healy, Coughlan, Williams, Clarke, Kelleher and Cannon2019b, Reference Healy, Coughlan, Clarke, Kelleher and Cannon2020, Reference Healy, Eaton, Cotter, Carter, Dhondt and Cannon2021; McMahon et al., Reference McMahon, Corcoran, Keeley, Clarke, Coughlan, Wasserman and Cannon2020).