Wednesday, March 11, 2020

Prevention of Psychosis—Advances in Detection, Prognosis, & Intervention; but evidence that favored any particular preventive intervention over another is currently of great uncertainty

Prevention of Psychosis—Advances in Detection, Prognosis, and Intervention. Paolo Fusar-Poli et al. JAMA Psychiatry, March 11, 2020. doi:10.1001/jamapsychiatry.2019.4779

Key Points
Question  What is the status of current clinical knowledge in the detection, prognosis, and interventions for individuals at risk of psychosis?

Findings  In this review of 42 meta-analyses encompassing 81 outcomes, detecting individuals at risk for psychosis required knowledge of their specific sociodemographic, clinical, functional, cognitive, and neurobiological characteristics, and predicting outcomes was achieved with good accuracy provided that assessment tools were used in clinical samples. Evidence for specific effective interventions for this patient population is currently insufficient.

Meaning  Findings of this review suggest that, although clinical research knowledge for psychosis prevention is substantial and detecting and formulating a prognosis in individuals at risk for psychosis are possible, further research is needed to identify specific effective interventions in individuals with sufficient risk enrichment.

Importance  Detection, prognosis, and indicated interventions in individuals at clinical high risk for psychosis (CHR-P) are key components of preventive psychiatry.

Objective  To provide a comprehensive, evidence-based systematic appraisal of the advancements and limitations of detection, prognosis, and interventions for CHR-P individuals and to formulate updated recommendations.

Evidence Review  Web of Science, Cochrane Central Register of Reviews, and Ovid/PsychINFO were searched for articles published from January 1, 2013, to June 30, 2019, to identify meta-analyses conducted in CHR-P individuals. MEDLINE was used to search the reference lists of retrieved articles. Data obtained from each article included first author, year of publication, topic investigated, type of publication, study design and number, sample size of CHR-P population and comparison group, type of comparison group, age and sex of CHR-P individuals, type of prognostic assessment, interventions, quality assessment (using AMSTAR [Assessing the Methodological Quality of Systematic Reviews]), and key findings with their effect sizes.

Findings  In total, 42 meta-analyses published in the past 6 years and encompassing 81 outcomes were included. For the detection component, CHR-P individuals were young (mean [SD] age, 20.6 [3.2] years), were more frequently male (58%), and predominantly presented with attenuated psychotic symptoms lasting for more than 1 year before their presentation at specialized services. CHR-P individuals accumulated several sociodemographic risk factors compared with control participants. Substance use (33% tobacco use and 27% cannabis use), comorbid mental disorders (41% with depressive disorders and 15% with anxiety disorders), suicidal ideation (66%), and self-harm (49%) were also frequently seen in CHR-P individuals. CHR-P individuals showed impairments in work (Cohen d = 0.57) or educational functioning (Cohen d = 0.21), social functioning (Cohen d = 1.25), and quality of life (Cohen d = 1.75). Several neurobiological and neurocognitive alterations were confirmed in this study. For the prognosis component, the prognostic accuracy of CHR-P instruments was good, provided they were used in clinical samples. Overall, risk of psychosis was 22% at 3 years, and the risk was the highest in the brief and limited intermittent psychotic symptoms subgroup (38%). Baseline severity of attenuated psychotic (Cohen d = 0.35) and negative symptoms (Cohen d = 0.39) as well as low functioning (Cohen d = 0.29) were associated with an increased risk of psychosis. Controlling risk enrichment and implementing sequential risk assessments can optimize prognostic accuracy. For the intervention component, no robust evidence yet exists to favor any indicated intervention over another (including needs-based interventions and control conditions) for preventing psychosis or ameliorating any other outcome in CHR-P individuals. However, because the uncertainty of this evidence is high, needs-based and psychological interventions should still be offered.

Conclusions and Relevance  This review confirmed recent substantial advancements in the detection and prognosis of CHR-P individuals while suggesting that effective indicated interventions need to be identified. This evidence suggests a need for specialized services to detect CHR-P individuals in primary and secondary care settings, to formulate a prognosis with validated psychometric instruments, and to offer needs-based and psychological interventions.


Detection, assessment, and intervention before the onset of a first episode of the disorder in individuals at clinical high risk for psychosis (CHR-P) have the potential to maximize the benefits of early interventions in psychosis.1,2 The CHR-P paradigm originated in Australia 25 years ago3 and has since gained enough traction to stimulate hundreds of research publications. These published studies have been summarized in evidence synthesis studies spanning different topics and have influenced several national4 and international5 clinical guidelines and diagnostic manuals (eg, DSM-56,7). Overall, CHR-P represents the most established preventive approach in clinical psychiatry; therefore, periodically reviewing its progress and limitations is essential.

The rapid developments of detection, prognostic, and intervention-focused knowledge in the CHR-P field have not yet been integrated into a comprehensive, evidence-based summary since a 2013 publication in JAMA Psychiatry.8 Produced by the European College of Neuropsychopharmacology Network on the Prevention of Mental Disorders and Mental Health Promotion,9 the present study aimed to provide the first umbrella review summarizing the most recent evidence in the CHR-P field. An additional objective was to provide evidence-based recommendations for the 3 core components that are necessary to implement the CHR-P paradigm in clinical practice: detection, prognosis, and intervention.10

The protocol of this study was registered in PROSPERO (registration No. CRD42019135880). This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA) reporting guideline11 and the Reporting Items for Practice Guidelines in Healthcare (RIGHT) statement12 (eTable 1 in the Supplement).

Search Strategy and Selection Criteria
A multistep literature search was performed for articles published between January 1, 2013, and June 30, 2019 (eMethods 1 in the Supplement). Web of Science, Cochrane Central Register of Reviews, and Ovid/PsychINFO were searched for meta-analyses conducted in CHR-P individuals, and MEDLINE was used to search the reference lists of retrieved articles. The literature search, study selection, and data extraction were conducted independently by 2 of us (G.S.d.P., P.F.-P.), and consensus was reached through discussion.

Studies included were (1) meta-analyses (pairwise or network; aggregate or individual participant data) published as original investigations, reviews, research letters, or gray literature without restriction on the topic investigated13; (2) conducted in CHR-P individuals (ie, individuals meeting ultra-high-risk and/or basic symptoms criteria) as established by validated psychometric instruments8 (eMethods 2 in the Supplement) without restriction on the type of comparison group; and (3) published in the past 6 years.

Studies excluded (1) were original studies, study protocols, systematic reviews without quantitative analyses, and other non-meta-analytical studies; (2) did not formally assess and selected participants with established CHR-P instruments; or (3) were abstracts and conference proceedings. Data obtained from each article included first author, year of publication, topic investigated, type of publication, study design and number, sample size of CHR-P population and comparison group, type of comparison group, age and sex of CHR-P individuals, type of prognostic assessment, interventions, quality assessment (using AMSTAR [Assessing the Methodological Quality of Systematic Reviews]), and key findings with their effect sizes.

To respect the hierarchy of the evidence (eMethods 3 in the Supplement), if 2 or more meta-analyses addressing the same topic were found, we gave preference to individual participant data meta-analyses over aggregate network meta-analyses and to network meta-analyses over pairwise meta-analyses. The most recent study was selected when the previous criteria did not apply. If, after applying the hierarchical criteria, 2 studies were similar, both were included.

Outcome Measures, Data Extraction, and Timing and Effect Measures
From each study, a predetermined set of outcome measures (eMethods 4 in the Supplement) was extracted. The results were then narratively reported in tables, clustered around 3 core domains: detection, prognosis, and intervention.

When feasible, effect size measures were estimated through Cohen d. Other effect size measures were converted to Cohen d.13 In case of meta-analyses reporting time-dependent risks or rates or descriptive data only, proportions (95% CIs) or means (SDs) were summarized.

Quality Assessment
The quality of the included meta-analyses was assessed with the AMSTAR tool.14 Details of the meta-analyses and items evaluated are found in eMethods 5 in the Supplement.

Standards for Guidelines Development
To develop the recommendations, we followed the US Preventive Services Task Force (USPSTF) grading system15 (eTable 2 in the Supplement), which is suited explicitly for preventive approaches and has received extensive validation in articles published in several journals, including JAMA.16-21 Guideline development followed the JAMA Clinical Guidelines Synopsis, reaching consensus across the multidisciplinary European College of Neuropsychopharmacology Network on the Prevention of Mental Disorders and Mental Health Promotion.9 The rationale for the recommendations was provided. Conflicts of interest were fully detailed.

The literature search yielded 886 citations, which were screened for eligibility, and 55 of them were considered. After checking the inclusion and exclusion criteria, we included 42 meta-analyses encompassing 81 outcomes in the final analysis (Figure 1; eTables 3 to 11 in the Supplement).


No meta-analysis focused on the basic symptoms criteria. Overall, 85% (95% CI, 79%-90%) of CHR-P individuals met the attenuated psychosis symptoms (APS) criteria,22 10% (95% CI, 6%-14%) met the brief limited intermittent psychotic symptoms (BLIPS) criteria,22 and 5% (95% CI, 3%-7%) met the genetic risk and deterioration (GRD) syndrome criteria.22 The mean (SD) age of CHR-P individuals across the included studies was 20.6 (3.2) years, with a range of 12 to 49 years.5,22-52 These individuals were predominantly male (58%)22-29,31,33,35-43,46-50,53,54 and had attenuated psychotic symptoms lasting for more than 1 year before their presentation to specialized services. Several studies included underage patients.5,22-30,32-35,39-50,52,55,56 No differences were observed across the APS, BLIPS, and GRD subgroups.22 However, the mean (SD) duration of untreated attenuated psychotic symptoms tended to be shorter in the BLIPS group (435.8 [456.4] days) compared with the GRD group (783.5 [798.6] days) and APS group (709.5 [518.5] days) (eTable 3 in the Supplement).

Genetic and Environmental Risk and Protective Factors
Individuals who met CHR-P criteria, compared with those who did not, were more likely to have olfactory dysfunction (Cohen d = 0.71),57 be physically inactive (Cohen d = 0.7), have obstetric complications (Cohen d = 0.62), be unemployed (Cohen d = 0.57), be single (Cohen d = 0.27), have a low educational level (Cohen d = 0.21), and be male (Cohen d = 0.18).55 Trauma, which encompassed childhood emotional abuse (Cohen d = 0.98),55 high perceived stress (Cohen d = 0.85),55 childhood physical neglect (Cohen d = 0.62),55 and being bullied (Cohen d = 0.62)56 (eTable 4 in the Supplement; Figure 2), was also more frequent (87% for overall trauma)23 and severe (Cohen d = 1.38)56 in CHR-P individuals compared with the control groups. No meta-analysis addressed the association between genetic factors and the CHR-P state.

Substance Use
A statistically significant association was found between the CHR-P state and tobacco use (Cohen d = 0.61).55 Altogether, 33% of CHR-P individuals smoked tobacco compared with 14% in the control groups.58 Those in the CHR-P group were also more likely to be current cannabis users than control participants (27% vs 17%).53 Current cannabis use disorder was associated with an increased risk of psychosis (Cohen d = 0.31), whereas lifetime cannabis use was not.24 Higher levels of unusual thought content (Cohen d = 0.27) and suspiciousness (Cohen d = 0.21) were found in CHR-P individuals who were cannabis users compared with non–cannabis users,53 but attenuated positive or negative psychotic symptoms did not differ between these 2 groups53 (eTable 5 in the Supplement).

Clinical Comorbidity
Depressive (41%) and anxiety (15%) disorders were frequent in the CHR-P state.25 Most CHR-P individuals presented with suicidal ideation (66%).26 The prevalence of self-harm was 49% and of suicide attempts was 18% in CHR-P individuals26 (eTable 6 in the Supplement).

Functioning and Quality of Life
CHR-P individuals had lower levels of adolescence (Cohen d = 0.96-1.03) and childhood (Cohen d = 1.0) functioning compared with control participants.55 Functional impairments in CHR-P individuals were as severe as impairments in other mental disorders and were more severe than in control participants (Cohen d = 3.01)27 but were less severe than in established psychosis (Cohen d = 0.34). The CHR-P status was also associated with significant social deficits (Cohen d = 1.25).55 Quality of life was worse in CHR-P individuals than in control individuals (Cohen d = 1.75),27 whereas no differences from individuals with psychosis27 were reported (eTable 7 in the Supplement).

Visual learning (Cohen d = 0.27), processing speed (Cohen d = 0.42), and verbal learning (Cohen d = 0.42)54 were impaired in CHR-P individuals compared with control participants. CHR-P individuals who later developed psychosis showed poorer cognitive functioning (Cohen d = 0.24-0.54)54 compared with those who did not develop psychosis. However, no evidence of cognitive decline was found from baseline to follow-up in CHR-P individuals at any time.28 Although social cognition was impaired in CHR-P individuals compared with control individuals (Cohen d = 0.48),30 theory of mind was less impaired than in participants with first-episode psychosis (Cohen d = 0.45).31 CHR-P individuals showed more metacognitive dysfunctions (Cohen d = 0.57-1.09) than control participants but were similar to those with established psychosis29 (eTable 8 in the Supplement).

Neuroimaging and Biochemistry
CHR-P individuals had decreased blood interleukin 1β (IL-1β) levels33 (Cohen d = 0.66), increased salivary cortisol levels (Cohen d = 0.59),32 and increased blood IL-633 (Cohen d = 0.31) compared with control groups. The thalamus was smaller in CHR-P individuals than in control participants (Cohen d = 0.60),36 whereas no significant differences in the pituitary volume were found.37 The right hippocampal volume (unlike the left one) was also significantly smaller in CHR-P individuals38 compared with control participants (Cohen d = 0.24).38 Levels of glutamate and glutamine (measured together) were higher in the medial frontal cortex of CHR-P individuals than in control participants (Cohen d = 0.26).34

Compared with control individuals, CHR-P individuals showed decreased activations in the right inferior parietal lobule and left medial frontal gyrus and increased activations in the left superior temporal gyrus and right superior frontal gyrus35 (eTable 9 in the Supplement). As for neurophysiological processes, the mismatch negativity amplitude was reduced in CHR-P individuals compared with control participants (Cohen d = 0.4)39 and in CHR-P individuals who developed psychosis compared with those who did not (Cohen d = 0.71).59 A theoretical neurobiological model of the CHR-P state, which integrates these findings, is reported in Figure 3.60

Overall Prognostic Performance
Currently used semistructured interviews for psychosis prediction have an excellent overall prognostic performance (area under the curve [AUC] = 0.9).42 However, their sensitivity is high (96%) and specificity is low (47%),42 and they are not valid outside clinical samples that have undergone risk enrichment (ie, screening the general population is not useful)42 (Figure 2). The CAARMS (Comprehensive Assessment of At-Risk Mental States) instrument has an acceptable (AUC = 0.79) accuracy for predicting psychosis,43 and it has no substantial differences in prognostic accuracy from other CHR-P instruments,42 although the Structured Interview of Psychosis-Risk Syndromes has a slightly higher sensitivity (95%) than the CAARMS (86%).43 The reason for this lack of difference in prognostic accuracy is that most of the risk for psychosis (posttest risk) is accounted for by the way these individuals are recruited and sampled (pretest risk, independent from clinically verified CHR-P status) before the CHR-P test is administered.41 Pretest risk for psychosis is 15% at 3 years and is heterogeneous, ranging from 9% to 24%. Variability in pretest risk for psychosis is modulated by the type of sampling strategies,41 increasing if samples are recruited from secondary care and decreasing if samples are recruited from the community41 (Figure 2; eTable 10 in the Supplement).

The proportion of CHR-P individuals who developed a psychotic disorder (positive posttest risk, updated in 2016) was 22% at 3 years (Figure 4).40 The speed of transition to psychosis was greatest in the first months after CHR-P individuals presented to clinical services (median time to psychosis = 8 months).61 Transition to schizophrenia-spectrum psychoses was more than 6 times more frequent (73%) than transition to affective psychoses (11%), whereas transition to other psychoses was 16%.40 The transition risk to psychosis was higher in the BLIPS subgroup (38%) than in the APS (24%) and GRD (8%) subgroups at the 48-month follow-up or later,22 whereas the GRD subgroup was not at higher risk compared with the help-seeking control participants (which represents the standard comparative group during CHR-P interviews).22 No prognostic difference in the risk of psychotic recurrence was found across different operationalizations of short-lived psychotic episodes, including acute and transient psychotic disorders and brief psychotic disorders, but this risk was lower than in patients with remitted first-episode schizophrenia62 (eTable 10 in the Supplement). In the BLIPS group, the 2-year risk of developing schizophrenia was 23% and affective psychoses was null.62 Conversely, the remission rate of the baseline CHR-P symptoms was 35% at 1.94 years’ follow-up.45 No data were available on the remission rates across the BLIPS, APS, and GRD subgroups.

Prediction of Outcomes
Among CHR-P individuals, transition to psychosis was associated with severity of negative symptoms (Cohen d = 0.39), right-handedness (Cohen d = 0.26), severity of attenuated positive psychotic symptoms (Cohen d = 0.35), disorganized and cognitive symptoms (Cohen d = 0.32), unemployment (Cohen d = 0.32), severity of total symptoms (Cohen d = 0.31), low functioning (Cohen d = 0.29), severity of general symptoms (Cohen d = 0.23), living alone (Cohen d = 0.16), male sex (Cohen d = 0.10), and lifetime stress or trauma (Cohen d = 0.08) (eTable 11 in the Supplement; Figure 2).63 However, only severity of attenuated psychotic symptoms and low functioning (highly suggestive level of evidence13) and negative symptoms (suggestive level of evidence13) were associated with psychosis onset after controlling for several biases.63 Comorbid anxiety and depressive disorders were not significantly associated with transition to psychosis.25 No data were available on the predictors of outcomes other than psychosis onset.

Prognostic accuracy may be optimized by controlling pretest risk enrichment55 and using sequential assessments that include a staged assessment based on clinical information, electroencephalogram, neuroimaging, and blood markers44 (eTable 11 in the Supplement).

No evidence was found that favored any indicated intervention over another (including needs-based interventions or control conditions) for preventing the transition to psychosis.46 Likewise, no evidence supported the superior efficacy of any intervention over another for reducing attenuated positive psychotic symptoms47,48 (2 meta-analyses on the same topic were retained after the hierarchical criteria were applied) or negative symptoms,49 improving overall functioning5 or social functioning,50 alleviating depression,52 improving symptom-related distress or quality of life,51 or affecting acceptability46 in CHR-P individuals (eTable 12 in the Supplement).

To our knowledge, this study is the first comprehensive review (42 meta-analyses with 81 outcomes) focusing on detection, prognosis, and intervention of CHR-P individuals. No meta-analyses had reported consistent results from well-designed, well-conducted studies related to detection, prognosis, or interventions in representative primary care populations (USPSTF criteria for high level of certainty).

The detection of CHR-P individuals has a moderate level of certainty (grade B; Table). Research in the past 6 years has revealed that detection of truly at-risk individuals may be the key rate-limiting step toward a successful implementation of the CHR-P paradigm at scale. Although the CHR-P group is heterogeneous, its baseline sociodemographic characteristics are now clearer; typically, these individuals were young (mean [SD] age, 20.6 [3.2] years) men (58%) who presented with APS and had associated impairments in global functioning (Cohen d = 3.01), social functioning (Cohen d = 1.25),55 and quality of life (Cohen d = 1.75)27; suicidal ideation (66%26); self-harm (49%26); and suicide attempts (18%26). Because of these problems, these individuals sought help at specialized clinics; however, typically, these problems remained undetected (and untreated) for 1 year or more.

Currently, detection of CHR-P individuals is entirely based on their referral on suspicion of psychosis risk and on the promotion of help-seeking behaviors. These detection strategies appear inefficient: only about 5%64 to 12%65 of first-episode cases were detected at the time of their CHR-P stage through stand-alone or youth mental health services. A further caveat is that approximately one-third of first-episode cases may not lead to the development of psychosis through a CHR-P stage.66,67 Furthermore, at presentation, CHR-P individuals often had comorbid nonpsychotic mental disorders (41% depressive disorders and 15% anxiety disorders25) and substance use (33% tobacco use58 and 27% cannabis use53). Because of these limitations, the chain of evidence lacked coherence (per USPSTF grading system; eTable 2 in the Supplement). These issues could be addressed by integrated detection programs that leverage automatic detection tools for screening large clinical10,64,68 and nonclinical69 samples in a transdiagnostic70 fashion, encompassing primary and secondary care, the community,71 and youth mental health services.72 In addition, the detection of CHR-P individuals is currently based on the assessment of symptoms, but symptoms may be only the epiphenomena of underlying pathophysiological processes. CHR-P individuals often have several established sociodemographic, environmental, and other types of risk factors for psychosis,73 including male sex, unemployment, single status, low educational and functional level, obstetric complications, physical inactivity, olfactory dysfunction, and childhood trauma (Figure 2; eDiscussion 1 in the Supplement). Incorporating the assessment of these multiple factors with CHR-P symptoms, resulting in a Psychosis Polyrisk Score, may produce refined detection approaches74 that better map the etiopathological path of psychosis onset.

The prognosis of CHR-P individuals has a moderate level of certainty (grade B; Table).47,75 Converging evidence has demonstrated that CHR-P assessment instruments have good prognostic accuracy (AUC = 0.9)42 for the prediction of psychosis, comparable to the accuracy of clinical tools used in other areas of medicine.42 However, alternative instruments are needed to predict other nonpsychotic outcomes (eg, bipolar onset in those at risk76,77). No substantial prognostic accuracy differences were found across different CHR-P tools.42 The current CHR-P prediction instruments have high sensitivity (96%) but low specificity (47%) and are valid only if applied to clinical samples that have accumulated the above risk factors and have therefore already undergone substantial risk enrichment (Figure 2). In fact, it is not only CHR-P criteria that determine the probability of transition to psychosis but also the recruitment and selection of samples, which modulate enrichment in risk.47,78 The next generation of research should better deconstruct and control risk enrichment79 to maximize the scalability of the use of the CHR-P prediction instruments.71 The 3-year meta-analytic risk of psychosis onset in the entire CHR-P group has declined from 31.5% (estimated in 201280) to the current 22% (Figure 4), although not globally.81

Transition risk has decreased when recruitment strategies focused on the community as opposed to primary or secondary care (eDiscussion 2 in the Supplement). Risk was the highest in the BLIPS subgroup (38% at 4 years; 89% at 5 years if there were “seriously disorganising or dangerous”82 features), intermediate in the APS subgroup (24% at 4 years), and lowest in the GRD subgroup (8% at 4 years).22 Those in the GRD subgroup were not at higher risk than the help-seeking control individuals at up to 4 years of follow-up.22 A revised version of the CHR-P paradigm, which includes stratification across these 3 subgroups, has therefore been proposed.2,83 The BLIPS group also overlapped substantially with the acute and transient psychotic disorders in the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision.82 Therefore, current CHR-P instruments can allow only subgroup-level (ie, BLIPS>APS>GRD) but not participant-level prognosis (inconsistent evidence, USPSTF grading; eTable 2 in the Supplement).

To refine prognosis at the individual participant level, future research may consider specific risk factors (eg, sex, stress and trauma, employment, and living status63), biomarkers (eg, hippocampal volume38), or cognitive markers (eg, processing speed, verbal and visual memory, and attention84) in addition to the CHR-P subgroups22 and clinical symptoms (only severity of attenuated positive and negative symptoms and level of functioning are robust risk factors for psychosis63). The potential of this approach has been supported by the development and validation of individualized clinical prediction models that leverage multimodal risk profiling,64,85,86 including dynamic87 risk prediction models.88 Because these models tend to be more complex compared with standard symptomatic CHR-P assessments, they are more likely to enter clinical routine through a sequential testing framework44 (eDiscussion 3 in the Supplement). Good outcomes in CHR-P individuals have not been fully operationalized,89 and information is lacking on prediction of relevant clinical outcomes (USPSTF; eTable 2 in the Supplement), such as functional level and quality of life, with only approximately one-third of individuals remitting from their initial CHR-P state.45

The available evidence is insufficient (grade C/I; Table) to assess the effects of preventive interventions on health outcomes in CHR-P groups. Although earlier meta-analyses found advantages to cognitive behavioral therapy,90 which is currently recommended by clinical guidelines,4 the inclusion of new trials in recent meta-analyses has indicated no clear benefits to favor any available intervention over another intervention or any control condition, such as needs-based interventions. An independent pairwise meta-analysis published by the Cochrane Group after completion of the present study concluded that no convincing, unbiased, high-quality evidence exists that favors any type of intervention.91 Evidence is insufficient because these studies tended to report large CIs and therefore high uncertainty (USPSTF; eTable 2 in the Supplement) in the meta-analytic estimates, and significant implications of the interventions for specific subgroups may not have been detected. For example, the needs-based interventions that are typically used as control conditions may have diluted the comparative efficacy of experimental interventions. This nondifferential outcome could also be an effect of the sampling biases leading to too few CHR-P individuals in the intervention studies who were at true risk for psychosis, diluting the statistical power of current trials that may have not been able to detect small to modest effect sizes (USPSTF; eTable 2 in the Supplement).92 This lack of demonstrable advantages of specific interventions could also be the consequence of one-size-fits-all approaches in treating CHR-P individuals that go against the clinical, neurobiological, and prognostic heterogeneity of this group and against the recent calls for precision medicine. For example, CHR-P interventions to date have been developed largely for individuals with APS at the expense of those with BLIPS, who are often unwilling to receive the recommended interventions. Another explanation for the lack of comparative efficacy of preventive interventions is that they have largely targeted symptoms, as opposed to key neurobiological processes associated with the onset of psychosis (gaps in the chain of evidence, USPSTF; eTable 2 in the Supplement; Figure 3) or risk factors that could be modified (eg, physical inactivity; Figure 2). We believe that future experimental interventions should also better target relevant outcomes (USPSTF; eTable 2 in the Supplement) other than psychosis onset, including functioning, given the poor remission rates and low functioning of this population.93 As acknowledged by the USPSTF criteria (eTable 2 in the Supplement), in the case of uncertainty, new trials published in the near future may allow a more accurate estimation of the preventive implications for health outcomes.

Grading the recent meta-analytic evidence described in this review, the European College of Neuropsychopharmacology Network on the Prevention of Mental Disorders and Mental Health Promotion has recommended (Table) implementing specialized services to detect CHR-P individuals in primary and secondary care settings and to formulate a prognosis with the validated psychometric instruments.9 Owing to insufficient evidence that favored any particular preventive intervention over another (including control conditions) and considering the uncertainty of the current evidence, no firm conclusions can be made,91 and a cautious approach is required. This approach should involve offering the least onerous feasible primary indicated prevention based on needs-based interventions and psychotherapy (cognitive behavioral therapy or integrated psychological interventions), titrated in accordance with the patient characteristics and risk profile (CHR-P subgroup levels BLIPS>APS>GRD, severity of attenuated positive and negative symptoms, and level of functioning), values, and preferences of the individual.94,95 In addition, other comorbid psychiatric conditions should be treated according to available guidelines, aiming for improving recovery, functional status, and quality of life beyond preventive aims.

The main limitations of this study were that the meta-analyses had heterogeneous quality (eResults in the Supplement) and that the literature search approach may have favored the selection of more commonly and readily studied domains that are more likely to be included in a meta-analysis. We cannot exclude the possibility that some promising advancements in the CHR-P field, despite having sufficient data, do not (yet) have a corresponding eligible meta-analysis, such as polygenic risk scores.96 However, in the current era, this possibility is becoming increasingly less likely, with meta-analyses being performed frequently, to the point that multiple meta-analyses are available for the same topic.97-99 In any case, for most putative domains that are difficult to study (or uncommonly studied), the current grade of evidence is unlikely to be remarkable, given the limited data.

Over recent years, substantial advancements in the detection and prognosis of CHR-P individuals have been confirmed in several meta-analyses. However, further research is needed to optimize risk enrichment and stratification and to identify effective interventions that target quantitative individualized risk signatures for both poor and good outcomes.

Longer Religious Fasting Increases Support for Islamist Parties: Evidence from Ramadan

Aksoy, Ozan, and Diego Gambetta. 2020. “Longer Religious Fasting Increases Support for Islamist Parties: Evidence from Ramadan.” SocArXiv. March 10. doi:10.31235/

Abstract: Much scientific research shows that the sacrifices imposed by religious practices are positively associated with the success of religious organizations. We present the first evidence that this association is causal. We employ a natural experiment that rests on a peculiar time-shifting feature of Ramadan that makes the length of fasting time vary from year-to-year and by latitude. We find that an hour increase in fasting during the median Ramadan day increases the vote shares of Islamist political parties by about 6.5 percentage-points in Turkey’s parliamentary elections between 1973 and 2018. This effect is weaker in provinces where the proportion of non-orthodox Muslims is higher, but stronger in provinces where the number of per capita mosques and of religious personnel is higher. Further analyses suggest that the main mechanism underlying our findings is an increased commitment to religion induced by costlier practice. By showing that the success of religious organizations is causally related to the sacrifice demanded by religious practices, these results strengthen a key finding of the science of religion.

Check also Witnessing fewer credible cultural cues of religious commitment is the most potent predictor of religious disbelief, β=0.28, followed distantly by reflective cognitive style:
Gervais, Will M., Maxine B. Najle, Sarah R. Schiavone, and Nava Caluori. 2019. “The Origins of Religious Disbelief: A Dual Inheritance Approach.” PsyArXiv. December 8.

Bullshitting frequency was negatively associated with sincerity, honesty, cognitive ability, open-minded cognition, and self-regard and positively related to overclaiming

Littrell, Shane, Evan F. Risko, and Jonathan A. Fugelsang. 2019. “The Bullshitting Frequency Scale: Development and Psychometric Properties.” PsyArXiv. September 27. doi:10.31234/

Abstract: Recent psychological research has identified important individual differences associated with receptivity to bullshit, which has greatly enhanced our understanding of the processes behind susceptibility to pseudo-profound or otherwise misleading information. However, the bulk of this research attention has focused on cognitive and dispositional factors related to bullshit (the product), while largely overlooking the influences behind bullshitting (the act). Here, we present results from four studies focusing on the construction and validation of a new, reliable scale measuring the frequency with which individuals engage in two types of bullshitting (persuasive and evasive) in everyday situations. Overall, bullshitting frequency was negatively associated with sincerity, honesty, cognitive ability, open-minded cognition, and self-regard and positively related to overclaiming. Additionally, the Bullshitting Frequency Scale was found to reliably measure constructs distinct from lying. These results represent an important step forward by demonstrating the utility of the Bullshitting Frequency Scale as well as highlighting certain individual differences that may play important roles in the extent to which individuals engage in everyday bullshitting

Does Honesty Require Time? Shalvi, Eldar, and Bereby-Meyer (2012) may have overestimated the true effect of time pressure on cheating and the generality of the effect beyond the original context

Does Honesty Require Time? Two Preregistered Direct Replications of Experiment 2 of Shalvi, Eldar, and Bereby-Meyer (2012). Ine Van der Cruyssen et al. Psychological Science, March 10, 2020.

Abstract: Shalvi, Eldar, and Bereby-Meyer (2012) found across two studies (N = 72 for each) that time pressure increased cheating. These findings suggest that dishonesty comes naturally, whereas honesty requires overcoming the initial tendency to cheat. Although the study’s results were statistically significant, a Bayesian reanalysis indicates that they had low evidential strength. In a direct replication attempt of Shalvi et al.’s Experiment 2, we found that time pressure did not increase cheating, N = 428, point biserial correlation (rpb) = .05, Bayes factor (BF)01 = 16.06. One important deviation from the original procedure, however, was the use of mass testing. In a second direct replication with small groups of participants, we found that time pressure also did not increase cheating, N = 297, rpb = .03, BF01 = 9.59. These findings indicate that the original study may have overestimated the true effect of time pressure on cheating and the generality of the effect beyond the original context.

Keywords: intuition, cheating, lying, honesty, replication, moral decision making, time pressure, open data, open materials, preregistered

What is people’s automatic tendency in a tempting situation? Shalvi et al. (2012) found that time pressure, a straightforward manipulation to spark “thinking fast” over “thinking slow,” provoked more cheating, and they concluded that people’s initial response is to serve their self-interest and cheat. We found no evidence that time pressure increased cheating in the die-roll paradigm. There are three possible reasons why replication studies do not produce the same results as the original study: (a) methodological problems in the replication study, (b) overestimation of the true effect size in the original study, or (c) differences between the studies that moderate the effect (Wicherts, 2018).
The first possibility is that methodological limitations in the replication study produced different results. In our first replication study, participants may not have fully appreciated the financial benefits of cheating. In our second replication study, relying on two test sites and offering the task in two languages may have increased error variance. But even for participants who performed the task in their native language, there was anecdotal support for the absence of a time-pressure effect (BF01 = 2.90).
The second possible explanation is that the original study overestimated the true effect size. The use of between-session rather than within-session randomization in the original study makes the experimenter aware of condition assignment and raises the possibility that the experimenter influenced the results (Rosenthal et al., 1963). Also, a single observation (in this case, a single reported die-roll outcome) per participant is likely to provide for a noisy measure. With low reliability, the results are more likely to vary per sample.
The third possible explanation is that the time-pressure effect on cheating is influenced by the context and that differences between the studies explain the different results. Our replications differed in several ways from the original, the most prominent being the country where the study was run, namely Israel in the original versus The Netherlands in the replications. The difference in test site raises the possibility of cross-cultural differences in intuitive dishonesty. Perceived country corruption, for instance, is related to the amount of cheating in the die-under-the-cup game (Gächter & Schulz, 2016). Then again, the large meta-analysis by Abeler, Nosenzo, & Raymond (2019) found that cheating behavior varies little by country. Still, it seems worthwhile to explore whether the automatic tendency to cheat may vary with culture.
In both our PDRs, people were predominantly honest, and we in fact found no evidence of cheating.4 Whereas Shalvi et al. (2012) originally reasoned that “time pressure evokes lying even in settings in which people typically refrain from lying” (p. 1268), our findings point to the possibility that the time-pressure effect is bound to settings that produce more pronounced cheating (e.g., when providing justifications for cheating).
In sum, our findings indicate that the original study by Shalvi et al. (2012) may have overestimated the true effect of time pressure on cheating or the generality of the effect beyond the original context. The vast majority of our participants were honest—even under time pressure. This finding casts doubt on whether people’s intuitive tendency is to cheat and fits better with a preference for honest behavior.

In the context of romantic attraction, beautification can increase assertiveness in women

In the context of romantic attraction, beautification can increase assertiveness in women. Khandis R. Blake, Robert Brooks, Lindsie C. Arthur, Thomas F. Denson. PLOS, March 10, 2020.

Abstract: Can beautification empower women to act assertively? Some women report that beautification is an agentic and assertive act, whereas others find beautification to be oppressive and disempowering. To disentangle these effects, in the context of romantic attraction we conducted the first experimental tests of beautification—on psychological and behavioral assertiveness. Experiment 1 (N = 145) utilized a between-subjects design in which women used their own clothing, make-up, and accessories to adjust their appearance as they normally would for a “hot date” (beautification condition) or a casual day at home with friends (control condition). We measured implicit, explicit, and behavioral assertiveness, as well as positive affect and sexual motivation. Experiment 2 (N = 40) sought to conceptually replicate Experiment 1 using a within-subject design and different measures of assertiveness. Women completed measures of explicit assertiveness and assertive behavioral intentions in three domains, in whatever clothing they were wearing that day then again after extensively beautifying their appearance. In Experiment 1, we found that women demonstrated higher psychological assertiveness after beautifying their appearance, and that high sexual motivation mediated the effect of beautification on assertive behavior. All effects were independent of positive affect. Experiment 2 partially replicated Experiment 1. These experiments provide novel insight into the effects of women’s appearance-enhancing behaviors on assertiveness by providing evidence that beautification may positively affect assertiveness in women under some circumstances.


Using a within-subjects design, we found that beautification increased explicit assertiveness to the extent that beautification increased women’s sexual motivation. For assertive consumer behavioral intentions, findings were mixed. Beautification had a direct effect on increasing willingness to endorse public consumer assertiveness, but the effect did not reach conventional levels of statistical significance. Beautification also elevated endorsement of private consumer assertiveness, but the effect was moderated by trait self-objectification. The more women tended to self-objectify, the more they reported willingness to engage in private consumer assertiveness after beautification. This effect was also moderated by sexual motivation, showing the same pattern. We found no effect of beautification, self-objectification, or sexual motivation on consumer assertiveness unrelated to appearance.

General discussion

Research derived from objectification theory has emphasized the negative consequences of beautification and related practices, highlighting that they harm women and are derived from a cultural context that disempowers them [1,57,24]. An alternative perspective, derived from sociometer theory, holds that beautification can benefit women by raising their self-esteem in important domains [8,11,12]. We added clarity to this research area by experimentally manipulating beautification through within- and between-subject designs, and subsequently measuring multiple indicators of assertiveness, as well as positive mood, sexual motivation, and self-objectification. Our results suggest that beautification can increase assertiveness in women, but that the effect may be domain-specific. These findings shed light on a key tension in female psychology by challenging the notion that beautification and related appearance-enhancing phenomena are necessarily disempowering.
Many of our effects were dependent on beautification increasing sexual motivation, with beautification elevating assertiveness only when it also elevated sexual motivation. This finding is consistent with previous research [30], and suggests that the effect of beautification on assertiveness depends upon the degree to which beautification increases the subjective feeling of sexual attractiveness. By including measures of behavioral assertiveness (Experiment 1) and assertive behavioral intentions in three domains (Experiment 2), we intended to distinguish whether beautification-induced assertiveness was domain-specific or domain-general. Unfortunately, results were inconclusive: We did not find a significant effect for beautification in our appearance-unrelated consumer assertiveness vignette, however, we did find that beautification increased assertive behavior in the mock job interview in Experiment 1 (to the extent that it also increased sexual motivation). Future work teasing out these effects would help to clarify the conditions under which beautification can increase assertiveness, and whether that increase is specific to the appearance domain, or whether effects might transfer to unrelated domains.
Beautification interacted with sexual motivation to increase explicit assertiveness in women, regardless of trait self-objectification. Surprisingly, trait self-objectification was positively associated with a beautification-induced willingness to act assertively in one of our vignettes. This finding is supportive of parallel work showing that self-objectification and its antecedents can raise women’s self-esteem in particular contexts [11,12]. Though these effects warrant replication, they suggest that conceiving of self-objectification as an entirely deleterious phenomenon may mischaracterize its psychological effects. The degree to which self-objectification may translate into enhanced female empowerment in some conditions is perplexing, yet it is also a worthwhile topic for future research.

Implications for understanding self-objectification

These results provide further insight into understanding women’s motivation for appearance- modifying behaviors, including self-objectification and self-sexualization. Many of these phenomena are motivated by desires to elevate attractiveness to new or existing romantic partners [38,39]. However, our findings suggest that women may also engage in these behaviors to increase assertiveness as well as mood. Thus, a desire for feeling empowered may partially account for women’s beautification practices and consumption of appearance-enhancing products. This conception offers a unique perspective on why women are more beauty-focused when the economy declines (the lipstick effect; [40]). Beautification may provide an affordable way to elevate the subjective experience of empowerment in ecological conditions that often constrain agentic action [41].
The negative effects of self-objectification—including usurping women’s attentional and cognitive resources and increasing the likelihood of mental health problems—usually result from intermediary processes, such as elevated body shame and body surveillance [3]. Our findings raise the possibility that beautification may not always elicit these intermediary processes, and our work suggests that beautification can elicit sexual motivation as well. A defining difference between whether beautification and related phenomena empower or disempower women, then, may depend upon which intermediary processes are elicited. For example, if beautification elicits appearance anxiety or body shame, it may reduce assertiveness; If beautification elicits sexual motivation or high self-esteem, it may heighten assertiveness. Future investigation into the intermediary processes induced by appearance-relevant behaviors on positive and negative psychological outcomes would be a welcome contribution to future work.
Contextual effects—such as the person a beautified woman believes is judging her [42]—are also likely to be important. We focused on beautification in one situation only, and it is unclear whether mandatory beautification in other contexts (e.g., stipulated by an employer for an important meeting) would show similar effects. Women often become targets of backlash when they act assertively, especially in domains that are stereotype-inconsistent [43], and attractive women may be especially likely to be targeted. Women who engage in beautification and appearance-enhancing phenomena can also become targets of aggression by others, men and women like [33,4447]. Thus, although increases in beautification may engender benefits to women, in certain contexts it may also engender costs. The contexts under which women may express assertiveness and beautify without suffering backlash effects, or the contexts under which women experience beautification as especially disempowering are important future research topics.

The paradox of sexualized beautification and female agency

Although the current work provides evidence for a conditional effects of beautification on female assertiveness, our findings appear to be inconsistent with work showing that men and women perceive that women in attractive, revealing clothing lack agency [33,48,49]. Why is it that people perceive that women in such clothing lack agency, whereas the women themselves may potentially feel and behave in a more assertive manner? Compelling evidence demonstrates that people derogate those who act counter to the status quo [50]. Perceptions that women who engage in beautification lack agency may thus function to penalize women who threaten notions of demure and passive femininity through asserting sexual power [43,51]. Perceiving that these women lack agency may also support male dominance by discrediting the agency that some women demonstrate via beautification.
Equating beautification or self- sexualization with low agency may also reflect the cultural suppression of female sexuality, an ever-present albeit culturally variable phenomenon that sanctions women’s sexual self-expression more heavily than men’s. Although the drivers of the cultural suppression of female sexuality remain controversial [5254], evidence supports the idea that competition between women can encourage them to suppress the sexuality and attractiveness-enhancing efforts of other women. Derogating such women as cultural dupes, who misunderstand female agency and how they are perceived by others, may thus function to reduce the occurrence of competition amongst women by elevating anxiety in potential competitors. Ultimately, such a process may function to diminish the threat of another woman’s physical and sexual attractiveness.
Perceptions that sexualized women lack agency may also function to motivate sexual approach in men. Evidence suggests that some men find cues of sexual vulnerability and low agency in women to be alluring [55]. From a functional perspective, perceiving low agency in such women may be attractive to men because it reduces the threat of rejection, female infidelity, and paternity uncertainty associated with female sexual agency. It is also plausible that low agency women are perceived as less likely to rebuff sexual advances and easier to monopolize [30]. For these reasons, men’s perceptions of low agency in women may be a cognitive bias that engenders sexual approach, akin to the robust bias men show to over-estimating women’s sexual intent [56,57]. Future work investigating these notions would provide valuable insight into the constancy of patriarchal culture over time and provide. Research could also clarify the paradoxical nature of men’s views of women’s agency, and women’s view of their own agency.

Limitations and future directions

We aimed to provide a rigorous test of the effects of beautification on assertiveness by employing explicit, implicit, and behavioral indicators of assertiveness, ecologically valid designs, and by testing the importance of theoretically relevant mechanisms and confounds (i.e., sexual motivation, positive affect). That being said, our findings are limited in several ways. Although patterns of variation in Experiment 2 were generally consistent with Experiment 1, two effects from Experiment 2 did not reach conventional levels of statistical significance. Likewise, in Experiment 2, we failed to replicate the direct effect of beautification on explicit assertiveness, finding instead that the effect was dependent on beautification eliciting sexual motivation. This latter finding highlights the importance of sexual motivation to the beautification–assertiveness link, but it weakens our ability to draw conclusions about the overall relationship between the two phenomena. Likewise, whether assertiveness effects are domain-general, or specific to appearance-relevant domains only, was unresolved by the current work. Based on sociometer theory, we speculate that beautification-induced assertiveness may be strongest in appearance-related domains, and weaker, albeit present in other domains.
A further limitation is that the instructions in the beautification condition were multi-faceted. The instructions informed women to dress for a night out where they might meet someone they were romantically interested in, a hot date, and a party. We emphasized “hot date” in the verbal instructions most frequently both before and during the experimental sessions, and parties are locations where young people commonly meet romantic partners. We did so because the aim of our study was to focus on beautification in the context of romantic relationships, and attractiveness in the domain of romantic relationships is a domain where women are especially likely to derive self-esteem [8,11]. The multi-faceted nature of these instructions; however, may have introduced unnecessary noise in our experimental manipulation, weakening our ability to detect effects.
Another limitation is that design differences between Experiments 1 and 2 may account for some variability in our findings. Experiment 1 occurred in the laboratory, meaning that participants were seen by the experimenter after they changed their clothing and makeup. In contrast, Experiment 2 occurred online, and participants completed the experimental session in their home. We utilized this design difference so participants in Experiment 2 had the choice of their entire wardrobe and all of their own beauty products at their disposal. Unfortunately, this distinction between public and private may have weakened findings in Experiment 2. It is possible that the element of being seen in public after one enhances their sexual appearance strengthens the effect of beautification on female assertiveness, resulting in stronger effects in public versus private settings. Such an interpretation would account for weaker effects in Experiment 2 compared to Experiment 1.
A final limitation is that we only controlled for one individual difference in our analyses. Although trait self-objectification was highly relevant, many other individual differences affect women’s willingness to beautify, self-objectify, and self-sexualize. For example, recent work indicates that ideological components related to higher order personal values are especially relevant [58]. Testing whether findings reported here are sensitive to these differences, and the individual differences predictive of beautification, would strengthen our conclusions.

Weather and suicide: Association between meteorological variables and suicidal behavior—a systematic qualitative review article

Weather and suicide: Association between meteorological variables and suicidal behavior—a systematic qualitative review article. Charlotte Pervilhac M.Sc.-Psych., Kyrill Schoilew, Hansjörg Znoj & Thomas J. Müller. Der Nervenarzt vol 91, pages 227–232(2020).

Background: The effects of current and expected future climate change on mental health outcomes are of increasing concern. In this context, the importance of meteorological factors on suicidal behavior is receiving growing attention in research.

Objective: Systematic review article with qualitative synthesis of the currently available literature, looking at the association between meteorological variables and attempted and completed suicide.

Material and methods: Criteria-based, systematic literature search according to the PRISMA criteria. Peer-reviewed original research studies were included without time limits.

Results and conclusion: A total of 99 studies were included and grouped according to the research analysis based on daily, weekly, monthly and annual data. The majority of the studies reported a statistical association with at least one meteorological variable. The most consistent positive correlation was shown between temperature and suicidal behavior. However, the results are not conclusive and in part contradictory. The reported studies differed distinctively in terms of study design. Meteorological parameters may be associated with suicidal behavior. Future research in this area is needed to provide further clarity. Despite existing knowledge gaps, the current findings may have implications for suicide prevention plans.