Received: 11 January 2021 Revised: 9 July 2021 Accepted: 10 July 2021 DOI: 10.1002/cpp.2651 COMPREHENSIVE REVIEW Family environmental factors in at-risk mental states for psychosis Lídia Hinojosa-Marqués1 | Tecelli Domínguez-Martínez2 | Neus Barrantes-Vidal1,3,4 1 Departament de Psicologia Clínica i de la Salut, Universitat Autònoma de Barcelona, Abstract Barcelona, Spain The family environment represents an important psychosocial factor that impacts 2 Global Mental Health Research Center, psychosis prognosis, but little is known about its effect on the at-risk stages of psy- Directorate of Epidemiological and Psychosocial Research, ‘Ramo n de la Fuente chosis. This study presents a comprehensive review and summarizes the state of the Muñiz’ National Institute of Psychiatry, art of study on the wide range of family factors related to family functioning in the Ministry of Health, Mexico City, Mexico 3 Departament de Salut Mental, Sant Pere At-Risk Mental State (ARMS) for psychosis, as well as family interventions in ARMS  Sanitària, Barcelona, Spain Claver - Fundacio individuals. Publications were retrieved by an extensive search on MEDLINE, Psy- 4 Centre for Biomedical Research Network on cINFO and SCOPUS (1990–2020). Expressed Emotion is the most studied variable in Mental Health (CIBERSAM), Instituto de Salud Carlos III, Madrid, Spain ARMS literature, but there is scarce evidence of the role of other significant family factors at the ARMS stage. Overall, high Expressed Emotion did not appear to be Correspondence Neus Barrantes-Vidal, Departament de reactive to ARMS patients' poor clinical status. However, initial evidence has Psicologia Clínica i de la Salut, Facultat de suggested that relatives' beliefs about the disorder may play a significant role, either Psicologia, Edifici B, Universitat Autònoma de Barcelona, 08193 Bellaterra (Cerdanyola del as mediators of these relationships or as predictors of Expressed Emotion. Available Vallès), Barcelona, Spain. literature yet to yield a consistent pattern of findings on the association between Email: neus.barrantes@uab.cat Expressed Emotion or other family functioning indicators and negative outcomes, but Funding information some longitudinal studies highlight the greater potential for the protective effects of Centro de Investigacion Biomédica en Red de Salud Mental (CIBERSAM), Instituto de Salud positive family environments at the ARMS stage. Family-based interventions have Carlos III; Institucio Catalana de Recerca i demonstrated benefits for both ARMS individuals and family dynamics. An increased Estudis Avançats (ICREA) Academia Award; Comissionat per a Universitats i Recerca of focus on the impact of the at-risk stage of illness on relatives' mental well-being is Generalitat de Catalunya, Grant/Award required to provide family support based on their needs and to clarify the mecha- Number: 2017SGR1612; Spanish Ministerio de Economía y Competitividad, Grant/Award nisms leading to dysfunctional family dynamics during the critical ARMS period. Number: PSI2017-87512-C2-01 KEYWORDS caregivers, Expressed Emotion, family functioning, family interventions, high risk for psychosis, review 1 | I N T RO DU CT I O N family members' emotional attitudes towards the patient, as measured by the level of Expressed Emotion (EE; Brown et al., 1972), have The family environment has been extensively studied as an influential received most of the attention in psychosocial research. High EE (char- psychosocial factor that impacts psychosis prognosis (Butzlaff & acterized by the presence of elevated levels of criticism, hostility and/or Hooley, 1998; Kavanagh, 1992; Schlosser et al., 2012). In this context, Emotional Over-Involvement [EOI]) has consistently shown to be a robust predictor of relapse in Schizophrenia (Cechnicki et al., 2013; Lídia Hinojosa-Marqués and Tecelli Domínguez-Martínez have contributed equally to this work and share the first authorship. Marom et al., 2005; O'Driscoll et al., 2019; Wearden et al., 2000). This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. © 2021 The Authors. Clinical Psychology & Psychotherapy published by John Wiley & Sons Ltd. Clin Psychol Psychother. 2021;1–31. wileyonlinelibrary.com/journal/cpp 1 2  ET AL. HINOJOSA-MARQUES Within the literature on caregivers of individuals with Schizophre- nia, the theoretical attributional model (Barrowclough et al., 1994; Bar- Key Practitioner Message rowclough & Hooley, 2003) was developed for elucidating the • Relatives' psychological needs in the at-risk for psychosis relationships between relatives' illness attributions about psychosis and period have been overlooked, and there is an urgent need EE. The attributional model postulates that relatives' beliefs about the for more research to better understand the impact of the causes of the patients' illness are linked to relatives' emotional attitudes pre-psychotic stage on relatives' mental well-being. towards patients. In this regard, critical relatives tend to blame patients • Longitudinal studies highlight the significant role of posi- for their behaviours because they believe that the patient can control tive family aspects as predictors of at-risk for psychosis their symptoms, whereas relatives who feel excessively blameworthy patients' clinical and/or functional improvement over time. regarding the patient's illness may resort to over-involvement or self- • The effect of criticism on ARMS patients' clinical course sacrificing attitudes (Bentsen et al., 1998). is inconsistent across studies, while emotional over- A parallel line of research focused on the impact of the psychotic involvement appears to act as a protective factor and disorder on the family well-being has demonstrated that the responsi- could have some positive effects on patients' outcomes bility of caring for a family member with a psychotic disorder can lead at the ARMS stage. to show elevated levels of distress, anxiety and depression in care- • Relatives' illness attributions seem to be an important pre- givers (Barrowclough et al., 1996; Jansen, Gleeson, & Cotton, 2015; dictor factor and mediator in the relationship between Jungbauer & Angermeyer, 2002). To date, most research on family expressed emotion and ARMS clinical/functional features. factors has mainly been carried out on patients with chronic forms of • Family-based interventions have demonstrated benefits the psychosis phenotype (for a review, see Butzlaff & Hooley, 1998; for both ARMS individuals and family dynamics. Caqueo-Urízar et al., 2014; Hooley, 2007; Miklowitz, 2004) or individ- uals with recent onset of psychosis (i.e., first episode of psychosis [FEP] patients) (for a review, see Jansen, Gleeson, & Cotton, 2015; Koutra et al., 2014). Conversely, the family environment has been less caring for an individual with psychosis are at-risk for difficulties in studied in the phases preceding the onset of illness (i.e., high risk for family functioning related to cohesion, flexibility and/or poor coping psychosis or At-Risk Mental State [ARMS] stage). styles (Friedmann et al., 1997; Gupta & Bowie, 2018; Phillips The ARMS stage comprises a heterogeneous group of symptoms et al., 1998; Raune et al., 2004). Furthermore, it appears that these traditionally described during the prodromal phase of psychosis. Cur- overall family functioning indicators may have an impact on psychiat- rent standard definitions and operationalization yield three clusters of ric outcomes (Gurak & Weisman de Mamani, 2016; Weisman, 2005). individuals: (1) people with attenuated positive psychotic symptoms; This parallel yet less explored line of research considers other family- (2) people who have experienced brief intermittent episodes of frank related constructs, which are of great relevance and warrant further psychotic symptoms lasting no more than a week with spontaneous discussion, especially in the earlier stages of psychosis. full recovery; and (3) individuals with either genetic risk (having a first- To date, only one systematic review has so far attempted to degree relative diagnosed with a psychotic disorder) or meeting the describe the effects of the EE construct during the at-risk for psycho- criteria for Schizotypal Personality Disorder in addition to a significant sis stage (Izon et al., 2018), but it is limited only to EE, without ana- decrease in social functioning in both cases (McGlashan et al., 2001; lysing its relationships with other significant family factors involved in Miller et al., 2003; Yung et al., 2004). family functioning that could explain the potential mechanisms under- The study of family factors at the ARMS stage has the advantage lying the manifestation of EE in at-risk stages. Therefore, the present of not being biased by the plethora of confounding factors associated study review seeks to expand Izon et al. (2018) by providing the state with chronic and/or more advanced stages of the illness (e.g., the of the art of study on the wide range of family environmental factors onset of marked clinical symptoms, impairment, major co-morbidity related to family functioning at the pre-psychotic stage and discuss and severe medication side effects) and could ultimately improve the evidence they provide in light of existing theories on EE develop- understanding of the early development of high-EE attitudes for pre- ment (Barrowclough et al., 1994; Barrowclough & Hooley, 2003; vention purposes. Furthermore, since family factors may play a critical Hooley, 2007; Kavanagh, 1992; Kuipers et al., 2006; Miklowitz, 2004; role in the management and outcome of psychosis, developing an Patterson, 2013; Patterson et al., 2000, 2005; Raune et al., 2004). understanding of the family functioning at the very early stages of Moreover, given the central role of the EE construct in this field, the psychosis can help develop early interventions designed to prevent examination of findings is followed by an in-depth analysis of the pos- negative dynamics in the family environment. sible explanatory models that could account for the emergence of EE Although the study of EE has played a central role in family in the earliest stage of psychosis. environment research, different theoretical frameworks of family The research reviewed is divided into four different sections: functioning have emphasized other-dimensional indicators of family welfare, such as levels of cohesion and adaptability within the 1. studies investigating the association of EE and other family family, quality of parent–child communication and/or problem-solving environment constructs with ARMS symptoms/functioning and/or skills. Importantly, findings from previous research show that relatives relatives' psychological variables;  ET AL. HINOJOSA-MARQUES 3 TABLE 1 Summary of studies researching family environment variables and family interventions in the At-Risk Mental Stage (ARMS) of psychosis Design and Study Location participants Family variables Family measures Key findings Studies investigating the association of EE and other family environment constructs with ARMS symptoms and/or functioning and/or relatives' psychological variables McFarlane and Portland, New Cross-sectional study EE SAS-III (relatives' - Criticism and EOI Cook (2007) Rochelle and 69 subjects with self-reports) were higher in Kingston (USA) established parents of patients psychotic disorders with established and their parents psychotic disorders 50 subjects with than in parents of prodromal ARMS individuals. symptoms and their - Warmth was parents significantly higher in ARMS families than in parents of patients with established psychotic disorders. - In ARMS families, maternal criticism and EOI increased over time, whereas maternal and parental warmth decreased over time after the first signs of illness began to appear. Meneghelli Milan (Italy) Cross-sectional study EE CFI (interview FEP families: et al. (2011) 77 FEP and their based on - High EE was related relatives relatives' to longer DUI. 66 ARMS and their reports) - High paternal EOI relatives related to DUP. - Patients' severity of symptoms/ functioning not related to EE. ARMS families: - High EE not related to DUI. - The severity of illness/functioning not related to EE. Smith et al. (2018) London (UK) Cross-sectional study EE CFI (interview - One third of the 80 FEP relatives and Subjective appraisals based on sample reported at ARMS relatives of caregiving relatives' least one incident of Perceived mental well- reports) patient-initiated being ECI (relatives' self- violence. reports) - Reports of violence SF-36 (relatives' were associated self-reports) with poorer mental well-being scores among caregivers and more negative appraisals of caregiving. - Patient-initiated violence also correlated with greater criticism and hostility expressed towards patients. (Continues) 4  ET AL. HINOJOSA-MARQUES TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings Hamaie et al. (2016) Sendai (Japan) Cross-sectional study EE FAS (relatives' FEP relatives: 43 FEP and their self-reports) - Criticism was relatives associated with 56 ARMS and their relatives' higher relatives depression symptoms and with FEP patients' negative and general symptoms. - Criticism was explained by relatives' depression symptoms and FEP patients' general symptoms in the regression analyses. ARMS relatives: - Criticism was unrelated to ARMS patients' clinical status. Criticism was explained by relatives' higher educational levels in the regression analyses. Hinojosa-Marqués Barcelona (Spain) Prospective study, EE FQ (relatives' self- - Relatives' et al. (2020) 6-month follow-up Perceptions of illness reports) psychological 55 ARMS and 36 FEP Distress IPQS-R (relatives' distress and illness patients and their self-reports) attributions relatives at baseline SCL-90-R predicted criticism 33 relatives of ARMS (relatives' self- and EOI over and and 13 relatives of reports) above patients' FEP patients at illness follow-up characteristics at baseline and 6-month follow-up. - Relatives' increased levels of anxiety, attributions of blame towards the patients, emotional negative representation about the disorder and decreased levels of self-blame attributions predicted EE criticism at baseline. - Relatives' anxiety and negative emotional representation of the disorder (i.e., negative emotional response to the patient's mental health problems) were the only  ET AL. HINOJOSA-MARQUES 5 TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings significant predictors of EE criticism at follow- up, whereas anxiety, attributions of control by the relative and an emotional negative representation about the disorder predicted EE-EOI both at baseline and in 6-month follow- up assessments. Domínguez- Barcelona (Spain) Cross-sectional study EE FQ (relatives' self- - Anxiety, depression Martínez 37 FEP relatives Perceptions of illness reports) and attributions of et al. (2017) 41 ARMS relatives Distress IPQS-R (relatives' blame towards the self-reports) patient predicted SCL-90-R relatives' criticism in (relatives' self- both ARMS and FEP reports) relatives' groups. - Attributions of control by the patient and emotional negative representation of the disorder predicted relatives' EOI. - Anxiety more strongly associated with relatives' criticism in ARMS relatives than in FEP relatives. - Anxiety associated with relatives' EOI in ARMS but not in FEP relatives. Hinojosa-Marqués, Barcelona (Spain) Cross-sectional study EE FQ (relatives' self- - Momentary criticism Domínguez- 34 ARMS relatives Domains of ESM reports) and EOI were Martínez, 21 FEP relatives questionnairea ESM significantly Cristo bal-Narváez, questionnaire associated with the et al. (2019) (relatives' self- two FQ-EE reports) dimensions, respectively, supporting the criterion validity of real-world assessed EE dimensions. - Momentary and FQ- EE dimensions were associated with decreased positive affect, as well as with appraisals of less effective coping in daily life. - Only momentary EE dimensions were (Continues) 6  ET AL. HINOJOSA-MARQUES TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings associated with increased momentary negative affect. - Momentary criticism and FQ-criticism were more consistently related to situational stress and burden than momentary EOI and FQ-EOI. - Neither momentary nor FQ-EE dimensions showed distinct patterns of associations with illness attributions. - Findings partly support the construct validity of momentary criticism and EOI as well as the construct and ecological validity of the FQ as a sensitive measure of EE. Hinojosa-Marqués, Barcelona (Spain) Cross-sectional study EE FQ (relatives' self- - Relatives' perceived Domínguez- 30 FEP relatives Perceived loss reports) loss was associated Martínez, 48 ARMS relatives Attachment MIV-TIG with EE dimensions. Sheinbaum, (relatives' self- - Relatives' attachment et al. (2019) reports) anxiety, but not PAM (relatives' avoidance, mediated self-reports) the relationship of perceived loss with both criticism and EOI. Tsai et al. (2015) Los Angeles (USA) Cross-sectional study EE PC and PW - Analyses examining 49 ARMS (patients' self- the entire sample Non-Latino White: 38 reports) showed that Latino: 11 patients' perceived levels of criticism were negatively associated with ARMS negative symptomatology. - Race/ethnicity moderated the relationship between perceived criticism/warmth and CHR clinical symptomatology. Carol and Boulder (USA) Cross-sectional study EE FMSS (interview - Relatives of ARMS Mittal (2015) 37 ARMS based on provided 42 control group relatives' significantly fewer reports) positive comments about patients than relatives of the control group.  ET AL. HINOJOSA-MARQUES 7 TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings - No significant differences emerged between relatives of ARMS and the control group in criticism or EOI. - Relatives' criticism was related at the trend level to ARMS patients' positive symptoms. - Relatives' EOI, warmth and positive comments were unrelated to ARMS patients' clinical status. Domínguez- Barcelona (Spain) Cross-sectional study EE FQ (relatives' self- - Relatives' EOI was Martínez 24 FEP and their Perceptions of illness reports) associated with et al. (2014) relatives IPQS-R (relatives' patients' higher 20 ARMS and their self-reports) negative and general relatives symptoms, and worse social and role functioning. - Relatives' criticism was associated with patients' higher positive, negative and general symptoms, depressive symptoms and worse social and role functioning. - The association between EE and patients' symptoms/ functioning did not differ between FEP and ARMS groups. - Attributions of blame towards the patient mediated most of the relationships between EE and illness severity. Welsh and Northeast England Cross-sectional study Family perceptions FPS (patients' - ARMS and FEP Tiffin (2015) (UK) 26 FEP self-reports) showed more 44 ARMS maladaptive family 140 control group perceptions than the control group. - ARMS showed poorer perceived problem-solving and nurturing behaviour than the control group. - Family perceptions not related to symptom severity in (Continues) 8  ET AL. HINOJOSA-MARQUES TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings either ARMS and FEP groups; however, higher scores on the EE subscale of the FPS were related to manic symptomatology in both ARMS and FEP groups. Bentley et al. (2016) Baltimore (USA) Cross-sectional study Quality of parent–child BASC-2 (patients' - ARMS reported less 36 ARMS relationship self-reports) positive parent– 60 help-seeking child relationships control group (P-CH-R) and a higher perception of social stress than the control group. - P-CH-R moderated the relationship between ARMS diagnosis and perception of social stress. - ARMS informing poorer P-CH-R reported increased perceptions of social stress. - ARMS informing positive P-CH-R reported decreased perceptions of social stress. Thompson Maryland (USA) Cross-sectional study Family perceptions FAD (ARMS' self- - Perceived family et al. (2019) 52 ARMS reports) functioning moderated the effect of ARMS patients' symptoms on social/role functioning. - For individuals who perceived lower levels of family functioning, symptoms were moderately associated with social and role functioning. - For individuals who perceived higher levels of family functioning, symptoms were not significantly associated with social/role functioning. - Positive symptoms and perceived family  ET AL. HINOJOSA-MARQUES 9 TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings functioning were not associated, suggesting that perceived family functioning did not directly impact symptom severity and vice versa. Peh et al. (2020) Singapore (Singapore) Cross-sectional study Parental bonding PBI (patients' self- - ARMS individuals 164 ARMS reports) were more likely to 510 controls report having affectionless- controlling mothers, significantly lower maternal and paternal care and higher maternal and paternal overprotection than controls. - Higher paternal overprotection was significantly associated with a range of worse clinical and functioning outcomes among ARMS individuals. Predictive value of EE and other family environment constructs on symptoms and/or functional ARMS outcomes O'Brien et al. (2006) Los Angeles (USA) Prospective study, EE CFI (interview - Relatives' criticism at 3-month follow-up based on baseline was not 26 ARMS and their relatives' associated with primary caregivers reports) changes in at baseline and symptoms or social follow-up functioning at follow-up. - Relatives' EOI at baseline was associated with improvement in negative symptoms and social functioning at follow-up. - Relatives' positive remarks at baseline were associated with improvements in negative and disorganized symptoms at follow- up. - Relatives' warmth at baseline was associated with improvements in social functioning at follow-up. (Continues) 10  ET AL. HINOJOSA-MARQUES TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings Schlosser Los Angeles (USA) Prospective study, EE CFI (interview - Patients' perceived et al. (2010) 6-month follow-up based on criticism at baseline 63 ARMS and their relatives' significantly relatives at baseline reports) predicted worsening and follow-up PC and PW of attenuated (patients' self- positive symptoms reports) at follow-up. FMPC and FMPW - Relatives' criticism (relatives' self- and hostility at reports) baseline (as measured by CFI) significantly predicted worsening of attenuated positive symptoms at follow-up. - Relatives' EOI (moderate levels) and relatives' warmth (as measured by CFI) interacted such that they jointly predicted improved functioning at follow-up. O'Brien et al. (2008) Los Angeles (USA) Prospective study, EE CFI (interview - Relatives' positive 4-month follow-up Supportive and based on remarks at baseline 40 ARMS and their conflict-engaging relatives' were associated primary caregivers behaviour reports) with a decrease in at baseline and FIT (interactional negative symptoms follow-up patterns) at follow-up. - Relatives' warmth at baseline predicted an improvement in social functioning at follow-up. - Relatives' behaviours exhibited during the interactional task were not predictive of ARMS patients' symptoms/ functioning at follow-up. O'Brien et al. (2009) Los Angeles (USA) Prospective study, Social problem-solving Problem-solving - Baseline ARMS' 6-month follow-up behaviour discussion skilful problem- 33 ARMS and their (interactional solving, constructive primary caregivers patterns) communication and at baseline FIT (interactional relatives' 27 ARMS and their patterns) constructive primary caregivers CFI (interview communication at follow-up based on were associated relatives' with ARMS' reports) enhanced social functioning at follow-up. - Baseline ARMS' conflictual communication was  ET AL. HINOJOSA-MARQUES 11 TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings associated with increased positive symptoms at follow- up. Wang et al. (2015) Shanghai (China) Prospective study, Perceptions of family FAD (ARMS' self- - ARMS reported 6-month follow-up functioning reports) worse family 32 ARMS from the Family cohesion FACES-II (ARMS' functioning than the general population Family adaptability self-reports) control group. at baseline - Positive perception 25 ARMS from the of problem-solving general population and affective at follow-up responsiveness from 256 control group parents at baseline predicted less severe positive and negative symptoms at follow-up. - Better family cohesion and adaptability were associated with decreased general symptoms at baseline. - Better family cohesion and adaptability at baseline were associated with decreased disorganized and general symptoms at follow-up. Haidl et al. (2018) Multicentric European Prospective study, EE LEE (patients' self- - Patients' perceived study 18-month follow-up reports) irritability was found 235 ARMS at baseline to be a predictor of 205 ARMS at follow- conversion for up ARMS into FEP. - The importance of this family environmental risk factor was further demonstrated by an improvement of risk estimation in the original European Prediction of Psychosis Study (EPOS) predictor model. Family environmental factors as the main outcome explored in ARMS families Wong et al. (2008) New York (USA) Cross-sectional study Burden FEIS (interview - Levels of both 12 FEP and their based on subjective and relatives relatives' objective burden 11 ARMS and their reports) were comparable relatives between ARMS and FEP relatives. Gerson et al. (2011) New York (USA) Cross-sectional study Coping CCOPE (relatives' - Families reported 12 FEP relatives self-reports) moderate use of (Continues) 12  ET AL. HINOJOSA-MARQUES TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings 11 ARMS relatives ‘approach’ coping and rare use of ‘avoidant’ coping strategies. Yee et al. (2020) Cross-sectional study Coping with family RSQ (patients' - ARMS and control 75 ARMS stress self-report) groups reported 79 healthy controls similar levels of engagement strategies. - ARMS were more likely to use disengagement strategies when coping with family stress than the healthy control group. - Engagement strategies predicted greater perceptions of the availability of advice support. - Disengagement strategies were associated with anxiety and depression symptoms and predicted lower perceived social support from the family and greater family strain. Santesteban-Echarri Toronto (Canada) Cross-sectional study Family functioning FACES-IV - Families in all groups et al. (2018) 52 ARMS with early (patients' self- fell within ranges mood or anxiety reports) that reflected symptoms healthy family 108 ARMS with functioning. subthreshold - Family satisfaction is psychotic symptoms lower in youth at 41 non-help-seeking risk for severe youth with risk mental illness who factors present with early 42 control group signs of mood, anxiety or subthreshold psychotic symptoms. Salinger et al. (2018) 8 research centres of Cross-sectional study Family communication Problem-solving - Mothers of youth at the NAPLS-2 58 ARMS and their discussion clinical high risk for consortium respective relatives (interactional psychosis showed 58 at risk for bipolar patterns) significantly more disorder patients conflictual and less and their respective constructive relatives communication during problem- solving discussions than did mothers of youth at clinical high risk for bipolar disorder.  ET AL. HINOJOSA-MARQUES 13 TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings Baron et al. (2019) San Francisco (USA) Qualitative cross- Associative stigma Semi-structured - Perceptions of stigma sectional study interviews and coping with 12 ARMS relatives stigma emerged as main domains according to parents' experiences. - Parents expressed the increased experience of the stigma associated with psychotic-like symptoms as compared to other psychiatric symptoms. - Parents attempted to combating stigma through public disclosure with adolescents' privacy and autonomy needs. - Parents generally found that participation in research is less stigmatizing than non-speciality care. - Stigma impacted treatment-seeking and participation in family groups. Izon et al. (2019) Manchester (UK) Qualitative cross- Barriers and Semi-structured - Key barriers included sectional study facilitators when interviews a lack of 14 ARMS relatives providing support to understanding of ARMS patients individual's symptoms/ behaviour and how to support them, as well as unproductive coping strategies impacting on their health or/and well- being. - Key facilitators included early access and involvement of services, access to social support and open communication between family/ carers and the individual. Izon et al. (2020) Manchester (UK) Qualitative Family/carers' Semi-structured - Over 12 months, prospective study experiences interviews four factors were 10 ARMS relatives supporting ARMS important for family/carers to (Continues) 14  ET AL. HINOJOSA-MARQUES TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings individuals over facilitate their caring 12 months role: (a) looking after their well-being; (b) accessing additional support from family intervention; (c) communicating openly with the individual; and (d) engaging with services for the individual. - All these aspects were important for improving family communication, meeting family/ carers' unmet needs and helping them feel more confident and less isolated in their role as carers. Family intervention studies O'Brien et al. (2014) Eight research centres Randomized control Family communication Problem-solving - ARMS and their in the NAPLS-2 trial discussion respective families consortium Prospective study, (interactional assigned to FFT- 6-month follow up patterns) CHR showed Family Focused greater Therapy for improvement from individuals at clinical baseline to follow- high risk (FFT-CHR) up in constructive with 18 sessions communication and over 6 months vs. greater decreases in three-session family conflictual psychoeducational behaviours than intervention those who 51 ARMS and their participated in a relatives at baseline brief 38 ARMS and their psychoeducational relatives at follow- intervention. up - FFT-CHR was found to be effective in improving active listening and calm communication and decreasing irritability, anger, complaints and criticism in both patients and family caregivers. Miklowitz Eight research centres Randomized control - - - ARMS participants et al. (2014) in the NAPLS-2 trial assigned to FFT- consortium Prospective study, CHR showed 6-month follow up. greater Family Focused improvement in Therapy for positive symptoms individuals at clinical from baseline to high risk (FFT-CHR) follow-up than with 18 sessions participants  ET AL. HINOJOSA-MARQUES 15 TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings over 6 months vs. assigned to a brief three family family psychoeducation psychoeducational sessions intervention (three FFT-CHR was sessions of family administered to psychoeducation). patients and their - Negative symptoms parents improved 129 ARMS at baseline independently of 102 ARMS at follow- psychosocial up treatments. - Changes in psychosocial functioning over time were age dependent: patients over 19 showed greater role improvement in FFT-CHR, whereas participants aged between 16 and 19 showed more role improvement in the brief psychoeducational intervention. O'Brien et al. (2015) Eight research centres Randomized control EE PC (patients' and - Perceived maternal in the NAPLS-2 trial mothers' self- criticism decreased consortium Prospective study, reports) from pre- (baseline) 6-month follow up to post-treatment Family Focused (6 months) for both Therapy for treatment groups individuals at clinical (FFT-CHR and brief high risk (FFT-CHR) psychoeducational with 18 sessions intervention), and over 6 months vs. these changes in three-session family criticism predicted psychoeducational decreases in intervention attenuated positive FFT-CHR was symptoms at administered to 12-month follow-up. patients and their - Decreases in parents mothers' reports of 90 ARMS and their criticism were mothers at baseline marginally 41 ARMS and their significant mothers at follow- predictors (p = .06, up two-tailed). McFarlane Six mental health Quasi-experimental Only patient's - - No differences in et al. (2015) agencies and multisite trial outcome data were transition between settings belonging Prospective study, 6-, reported groups were found. to the Early 12- and 24-month - The FACT showed Detection and follow-up some beneficial Intervention for the Family-Aided Assertive effects over Prevention of Community community care in Psychosis Program Treatment (FACT) reduction of (EDIPPP) (United with a minimum symptoms and States) duration of 1 year improved global vs. community care functioning. (Continues) 16  ET AL. HINOJOSA-MARQUES TABLE 1 (Continued) Design and Study Location participants Family variables Family measures Key findings Interventions were - Over the 24-month administered to follow-up period, young people at risk the FACT group for psychosis and increased its level of their families participation in work 337 ARMS and their and school. relatives at baseline - Effects for 222 ARMS and their symptoms, relatives at functioning and 24-month follow-up global outcomes proved to be larger for the group having very early psychosis than for the clinical high-risk group. Landa et al., 2016 Open uncontrolled GF-CBT showed pilot trial significant decreases Prospective study, in symptoms and 3-month follow-up improvement in the Group- and Family- functioning of based Cognitive ARMS individuals. Behavioural Therapy Family members (GF-CBT) for youth showed significant at risk for psychosis improvements in the for 15 weeks use of CBT skills, 6 ARMS individuals enhanced and family members communication with at baseline their offspring and 5 ARMS individuals greater confidence and 4 family in their ability to members at help. 3-month follow-up Law et al. (2021) Two mental health Single (rater), blinded, FMSS (interview - The study recruited trusts in Northwest randomized based on 92% of the target, England (UK) feasibility study relatives' showing that it is Cross-sectional study reports) feasible to identify Combined Individual FQ (relatives' self- and recruit and Family reports) participants. Cognitive FMPC and FMPW - ARMS individuals Behavioural Therapy (relatives' self- presented with high (IFCBT) plus reports) levels of co-morbid enhanced treatment BDI (relatives' emotional disorders as usual (ETAU) vs. self-reports) such as moderate to ETAU alone for SIAS (relatives' severe depression ARMS and their self-reports) and high levels of family/carers ECR-RS (relatives' social anxiety. 70 ARMS and their self-reports) - CAARMS scores family/carers revealed high levels of distress in relation to symptoms, particularly on the non-bizarre ideas' subscale. - Family members/ carers also presented with moderate depression, anxiety and physical health problems.  ET AL. HINOJOSA-MARQUES 17 Note: The order of appearance of the studies follows the comments on the results in the body text to facilitate reading. Abbreviations: ARMS, At-Risk Mental State; BASC-2, Behaviour Assessment System for Children, Second Edition; BDI, Beck Depression Inventory; CCOPE, Carver's Coping Orientations to Problems Experienced Questionnaire; CFI, Camberwell Family Interview; DUI, duration of untreated illness; DUP, duration of untreated psychosis; ECI, Experience of Caregiving Inventory; ECR-RS, Relationship Structure Questionnaire; EE, Expressed Emotion; EOI, Emotional Over-Involvement; FACES-II, Family Adaptability and Cohesion Evaluation Scale II; FACES-IV, Family Adaptability and Cohesion Evaluation Scale IV; FAD, Family Assessment Device; FAS, Family Attitude Scale; FEIS, Family Experiences Interview Schedule; FEP, first episode of psychosis; FIT, The Family Interaction Task; FMPC, family member's perceived criticism; FMPW, family member's perceived warmth; FMSS, The Five Minute Speech Sample; FPS, Family Perception Scale; FQ, Family Questionnaire; IPQS-R, Illness Perception Questionnaire for Schizophrenia Relatives' version; LEE, Level of Expressed Emotion Scale; MIV-TIG, Mental Illness Version of the Texas Inventory of Grief; NAPLS-2 consortium, Emory University, Harvard University, University of Calgary, University of California Los Angeles, University of California San Diego, University of North Carolina, Yale University and Zucker Hillside Hospital; NAPLS-2, North American Prodrome Longitudinal Study; PAM, Psychosis Attachment Measure; PBI, Parental Bonding Instrument; PC, Perceived Criticism Scale; PW, Perceived Warmth Scale; RSQ, Responses to Stress Questionnaire; SAS-III, Social Adjustment Scale-III; SCL-90-R, Symptom Checklist-90-Revised; SF-36, The Short Form-36 Health Survey; SIAS, Social Interaction Anxiety Scale. a Momentary EE, affect at the moment, appraisals of effective coping, appraisals of the current situation, appraisals of burden, illness attributions, appraisals related to the self, positive appraisals of the patient and appraisals of the patient when relatives were in direct contact and/or had had recent contact with the patient. 2. the predictive value of EE and other family environment constructs recruited other relatives, such as stepparents, siblings and partners. on ARMS symptoms and/or functional outcomes; Additionally, we also included studies assessing family factors in 3. family environmental factors (other than EE) explored in ARMS mixed samples of ARMS patients and/or their family members with families; and healthy controls, FEP and/or clinical psychosis individuals given that 4. the effect of family interventions on symptomatic and/or mixed samples are common in the early psychosis literature. functional ARMS outcomes. 3 | RE SU LT S 2 | METHOD A total of 35 studies met the inclusion criteria. Table 1 summarizes 2.1 | Search strategy the main characteristics of the studies included. Publications were retrieved through an extensive search of three elec- tronic databases: MEDLINE, PsycINFO and SCOPUS (1990–2020). 3.1 | Study aims The search was conducted using the terms ‘expressed emotion’, ‘family expressed emotion’, ‘emotional over-involvement’, ‘criticism’, The aims and scope of the reviewed studies varied greatly. Regarding ‘hostility’, ‘warmth’, ‘family environment’, ‘family functioning’, ‘family the family environment variables, 15 studies (42.9%) explored EE cohesion’, ‘family adaptability’, ‘family burden’, ‘family coping’, ‘fam- (Carol & Mittal, 2015; Domínguez-Martínez et al., 2014, 2017; Haidl ily distress’, ‘relatives' illness attributions’, ‘caregivers’, ‘carers’, ‘rela- et al., 2018; Hamaie et al., 2016; Hinojosa-Marqués et al., 2020; tives’, ‘at-risk mental states’, ‘prodrome’, ‘ultra-high risk’, ‘clinical Hinojosa-Marqués, Domínguez-Martínez, bal-Narváez, Cristo high-risk for psychosis’ and ‘at-risk for psychosis’. et al., 2019; Hinojosa-Marqués, Domínguez-Martínez, Sheinbaum, et al., 2019; McFarlane & Cook, 2007; Meneghelli et al., 2011; O'Brien et al., 2006, 2008; Schlosser et al., 2010; Smith et al., 2018; Tsai 2.2 | Inclusion/exclusion criteria et al., 2015) and 14 (40%) explored other relevant family environment constructs as the main outcome (such as burden, problem-solving Inclusion criteria: (1) published in peer-reviewed journals in English; skills, communication, cohesion, adaptability, perception of the quality (2) date of publication from 1990 to 2020 (December); (3) samples of of relationships within the family, family functioning, coping, barriers ARMS patients and/or their relatives; and (4) research focused on and facilitators in the caring role) (Baron et al., 2019; Bentley (a) the impact of illness on family members' attitudes or psychological et al., 2016; Gerson et al., 2011; Izon et al., 2019, 2020; O'Brien health; (b) the influence of family environment variables on ARMS' et al., 2009; Peh et al., 2020; Salinger et al., 2018; Santesteban-Echarri clinical and/or functional outcomes; (c) relatives' psychological factors et al., 2018; Thompson et al., 2019; Wang et al., 2015; Welsh & accounting for EE; (d) the effects of family interventions on both Tiffin, 2015; Wong et al., 2008; Yee et al., 2020). Finally, six studies ARMS' clinical and/or functional outcomes; and (e) studies with at (17.1%) focused on family-based interventions and their effects on least one family measurement and/or relevant constructs to under- ARMS clinical outcomes, high-EE attitudes and/or family communica- stand the early emergence of EE. tion (Landa et al., 2016; Law et al., 2021; McFarlane et al., 2015; Exclusion criteria: (1) secondary publications (commentaries, edito- Miklowitz et al., 2014; O'Brien et al., 2014, 2015). Reviewed studies rials and letters); (2) theses, dissertations and conference papers; and also varied in their assessment methods. While some of them (3) case reports. We did not limit the search to studies that only assessed the family's emotional climate through relatives' self-reports included biological mothers and fathers but also included studies that and/or interviews based on relatives' reports (Carol & Mittal, 2015; 18  ET AL. HINOJOSA-MARQUES Domínguez-Martínez et al., 2014, 2017; Gerson et al., 2011; Hamaie patients have the same prevalence of high EE mostly due to EOI and et al., 2016; Hinojosa-Marqués et al., 2020; Hinojosa-Marqués, that EE was not related to the severity of symptoms or psychosocial bal-Narváez, et al., 2019; Hinojosa- Domínguez-Martínez, Cristo functioning in either group. The authors concluded that results par- Marqués, Domínguez-Martínez, Sheinbaum, et al., 2019; McFarlane & tially support the assumption that high-EE attitudes develop as a reac- Cook, 2007; Meneghelli et al., 2011; O'Brien et al., 2006, 2008, 2009; tion to patients' clinical/functional status. This led them to speculate Smith et al., 2018; Wong et al., 2008), other studies assessed family the existence of intrinsic components regarding family functioning variables from the patient's viewpoint (Bentley et al., 2016; Haidl which might also be involved in the genesis of EE (e.g., relatives' et al., 2018; Peh et al., 2020; Santesteban-Echarri et al., 2018; beliefs about the illness). In a slightly different way, Smith et al. (2018) Thompson et al., 2019; Tsai et al., 2015; Wang et al., 2015; Welsh & set out to examine how other patient-related variables (not strictly Tiffin, 2015; Yee et al., 2020), only two studies considered both clinical), such as patient-initiated violence, impacted the manifestation patients' and relatives' perspectives (O'Brien et al., 2015; Schlosser of EE, as well as on relatives' negative appraisals of caregiving and et al., 2010) and three very recent studies used a qualitative method- perceived mental well-being. They found that relatives' reports of vio- ology approach (Baron et al., 2019; Izon et al., 2019, 2020). Family lence in patients were associated with poorer mental well-being, more interaction tasks were also used to assess family-related variables negative appraisals of caregiving and greater criticism and hostility (O'Brien et al., 2008, 2009, 2014; Salinger et al., 2018). expressed towards patients. In a recent Japanese study, Hamaie et al. (2016) assessed the impact of patients' clinical characteristics on relatives' criticism by also 3.2 | Studies investigating the association of EE considering the contribution of relatives' psychological factors, in a and other family environment constructs with ARMS sample of 56 ARMS and 43 FEP patients with their caregivers. Find- symptoms/functioning and/or relatives' psychological ings showed that ARMS caregivers' criticism was not related to variables patients' clinical status or caregivers' depressive symptomatology but rather to caregivers' higher educational levels. Authors suggested that The scant existing literature on the relationship between EE and spe- high levels of education may contribute to developing higher expecta- cific symptoms and functioning in ARMS patients contains a substan- tions regarding the capacity for control of ARMS individuals over their tial level of contradictory findings, much more so than the research illness-related behaviours, which in turn would be expressed in critical conducted on patients with chronic Schizophrenia. attitudes. In a more comprehensive predictive model of EE, Hinojosa- Some cross-sectional studies have aimed to explore the assump- Marqués et al. (2020) also examined the contribution of early tion that EE develops as a reaction to patients' symptom severity by psychosis clinical and functional status to EE along with relatives' psy- comparing the prevalence of EE at different stages of illness severity chological factors. They explored the association of relatives' EE and/or by using retrospective reports on the duration of patients' dimensions (criticism and EOI) with relatives' psychological distress symptoms to examine their impact on EE (McFarlane & Cook, 2007; and illness attributions and with patients' clinical/functioning features Meneghelli et al., 2011). McFarlane and Cook (2007) did not use at baseline and 6-month follow-up in 91 dyads of early psychosis patients' measures to test their association with EE and instead com- patients and their relatives (comprising 55 ARMS and 36 FEP pared EE levels between families of patients with clinical psychosis patients). Furthermore, they explored whether relatives' psychological and ARMS individuals. Furthermore, they analysed the impact of rela- factors predicted EE dimensions over and above patients' baseline tives' perceived duration of prodromal symptoms on EE. They found clinical and functional status at both time points. The most relevant that parents of individuals with established psychotic disorders finding of this study revealed that relatives' psychological distress and reported higher levels of criticism and EOI and lower levels of warmth subjective appraisals of the illness accounted for significant variance than parents of ARMS individuals. Further analyses conducted only over and above patients' clinical and functional status in the predic- within the ARMS sample revealed that levels of maternal criticism and tion of criticism and EOI both at baseline and at 6-month follow-up EOI increased once the first signs of illness began to appear, whereas (please see Table 1 for more details of specific results). maternal and paternal warmth decreased over the pre-psychotic Other studies have placed greater emphasis on the relevance of development. The latter analyses were conducted using retrospective examining relatives' psychological variables to identify the mecha- reports from parents regarding the duration of the subpsychotic syn- nisms underlying EE. Specifically, Domínguez-Martínez et al. (2017) drome. This in turn provided a measure of parents' perceived duration highlighted the role of relatives' distress and cognitive representations of the prodromal symptoms, which was analysed in association with of illness. Results from this study showed that relatives' distress and the current EE values. The authors suggested that EE is a reaction to attributions of blaming patients for their illness predicted criticism in patients' deterioration, that is, increasing symptom severity and dis- both ARMS and FEP caregivers, whereas beliefs that symptoms are ability across the ARMS stage. Conversely, Meneghelli et al. (2011) within patients' control and an emotional negative representation stated that the duration of untreated illness was not related to EE in about the illness predicted EOI in both groups. On a correlational relatives of ARMS patients. Both duration of untreated illness and level, relatives' EE dimensions (i.e., criticism and EOI) were highly duration of untreated psychosis were measured from the relatives' associated with relatives' distress and several types of illness attribu- perspective. Results also showed that families of FEP and ARMS tions. Also, comparisons between groups revealed that ARMS  ET AL. HINOJOSA-MARQUES 19 relatives scored higher on criticism than FEP relatives. Moreover, anx- results showed cross-cultural differences among ARMS patients, iety was more strongly related to criticism in ARMS than FEP relatives the study has several limitations that do not allow to generalization of and associated with EOI in ARMS but not FEP relatives. the results.  bal- Findings from Hinojosa-Marqués, Domínguez-Martínez, Cristo Domínguez-Martínez et al. (2014) showed that relatives' criti- Narváez, et al. (2019) also highlighted the fact that EE dimensions (criti- cism and EOI were related to increased several symptom dimen- cism and EOI) were significantly associated with relatives' negative sions and functional impairment at both the ARMS and FEP stages affective states and negative illness attributions in the realm of daily life of psychosis. No differences emerged between ARMS and FEP rela- by using the Experience Sampling Method (ESM) in relatives of ARMS tives in terms of EE or its relationship with patients' symptoms/ and FEP patients. The authors examined the way EE dimensions, as functioning, suggesting that both patients' samples were similarly measured by momentary and psychometric self-reports (Family Ques- influenced by family environment variables. Going a step further, tionnaire [FQ]; Wiedemann et al., 2002), were expressed in daily life this study analysed the potential role of relatives' attributions in the and related to a wide variety of real-world experiences. Overall, these association between EE and patients' symptoms. It was found that findings provide a valid ecological insight of EE correlates, thus pointing relatives' attributions of blaming patients for their illness-related to the role of negative affective states and negative illness attributions difficulties mediated most of the relationships between EE and as potential mechanisms underlying the expression of EE (please see illness severity. Table 1 for more details of specific findings). Further research Some studies have examined how other family variables, as (Hinojosa-Marqués, Domínguez-Martínez, Sheinbaum, et al., 2019) rated by ARMS patients' perceptions, are related to patients' symp- underscored the importance of investigating the role of relatives' per- tom expression in the subclinical stages of psychosis. Welsh and ceived loss in the development of EE in early psychosis caregivers Tiffin (2015) showed that ARMS and FEP patients scored signifi- (i.e., ARMS and FEP relatives). No differences emerged between ARMS cantly higher on EE in comparison with a community control sample. and FEP relatives in terms of perceived loss. Findings indicated that rel- Interestingly, ARMS patients were likely to report poorer perceived atives' perceived loss was associated with both EE dimensions and that problem-solving and lower levels of nurturing behaviour in their relatives' attachment anxiety, but not attachment avoidance, mediated families compared with the community control group. Self-reported the relationship between perceived loss and EE components. These family perceptions were not related to symptom severity in either findings therefore also highlighted the importance of examining the role ARMS or FEP groups, although higher scores on the EE subscale of of relatives' attachment characteristics for understanding how percep- the self-reported family perceptions measure were significantly tions of loss could impact the manifestation of EE attitudes at the early related to manic symptomatology in both ARMS and FEP groups. stages of psychosis. Bentley et al. (2016) found that ARMS individuals reported less pos- Alternative cross-sectional studies focused mainly on examining itive parent–child relationships and higher levels of social stress than the influence of EE on ARMS patients' symptoms and/or functioning those who did not meet ARMS criteria but receive mental health (Domínguez-Martínez et al., 2014; Tsai et al., 2015). Carol and services. Parent–child relationships moderated the relationship Mittal (2015) have a comparable goal but also explored the relationship between ARMS diagnosis and social stress. The severity of social between EE and ARMS endocrine activity and self-concept. Findings stress was only dependent on the quality of family relationships in showed that relatives of ARMS reported similar levels of criticism and the ARMS group. ARMS individuals who reported poorer parent– EOI but significantly lower frequency of positive comments about child relationships tended to report greater perceptions of social patients in comparison with relatives of healthy individuals. Analyses of stress, whereas those who described positive parent–child relation- the ARMS group suggested that relatives' criticism (but not EOI) was ships tended to report lower perceptions of social stress. Similarly, related at the trend level to ARMS patients' positive symptoms. Neither Thompson et al. (2019) showed that ARMS individuals' perceived relatives' warmth nor positive comments were related to patients' family functioning moderated the effect of symptoms on social/role symptomatology. Results also revealed a moderate correlation at the functioning. Specifically, in patients who perceived lower levels of trend level between relatives' criticism and ARMS' heightened levels of family functioning, psychosis-risk symptoms were moderately associ- resting cortisol. Moreover, relatives' criticism was related to ARMS ated with social and role functioning. However, this effect was not increased negative self-concept while relatives' warmth was associated present in patients reporting higher levels of family functioning. A with ARMS increased positive belief about the self. recent study (Peh et al., 2020) explored parent–child relationships in Tsai et al. (2015) found that higher levels of patients' perceived ARMS individuals (n = 164), compared to controls (n = 510), in an maternal criticism were associated with lower levels of negative Asian setting by exploring the association between parental bond- symptoms across the entire sample of ARMS young adults. Authors ing, symptom severity and functioning. Findings showed that, com- speculated that criticism could also reflect mothers' involvement or pared with controls, ARMS individuals were more likely to report engagement in their offspring's care, thus serving as a protective fac- having affectionless, controlling mothers, significantly lower maternal tor in adolescent populations. Further analysis indicated that race/ and paternal care and higher maternal and paternal overprotection. ethnicity moderated the relationship between patients' perceived Moreover, higher paternal overprotection was significantly associ- levels of parental warmth/criticism and patients' symptom expression ated with a range of worse clinical and functioning outcomes among (please see Table 1 for more details of specific findings). Although the ARMS individuals. 20  ET AL. HINOJOSA-MARQUES 3.3 | Predictive value of EE and other family et al. (2018) followed up a large cohort of 235 ARMS individuals dur- environment constructs on ARMS symptoms and/or ing 18 months. Results revealed that ‘perceived irritability’, a domain functional outcomes of the Level of Expressed Emotion Scale (LEE; Cole & Kazarian, 1988), was found to be predictive for the conversion to psychosis. That is, Preliminary research examining family environment as a predictor of patients' perception of irritability (responsive to stress and less able to outcome in ARMS patients underlined the important role of a positive cope with it) by a key relative was a predictor of conversion from family environment. In a longitudinal study of 26 ARMS adolescents high-risk status to FEP. The predictive value of this family environ- and their primary caregivers, O'Brien et al. (2006) indicated that care- mental risk factor was further underpinned by an improvement of risk givers' criticism (assessed by the Camberwell Family Interview [CFI; estimation in the original EPOS psychosis prediction model Vaughn & Leff, 1976]) at baseline was unrelated to changes in symp- (Ruhrmann et al., 2010) by improving risk prediction above 0.90. toms or social functioning at 3-month follow-up. However, caregivers' EOI at baseline was associated with improvements in negative symp- toms and social functioning at follow-up. Likewise, caregivers' positive 3.4 | Family environmental factors (other than EE) remarks were associated with an improvement in negative and disorga- explored in ARMS families nized symptoms at follow-up, and caregivers' warmth was also associ- ated with enhanced social functioning at follow-up. Authors suggested Families' psycho-emotional and/or coping factors have scarcely been that EOI could be developmentally appropriate in adolescence, acting explored as the primary outcome (regardless of relatives' EE levels or as a protective factor. Along the same lines, Schlosser et al. (2010) the course of the illness in the affected relative) in the at-risk for suggested that when EOI is reported at moderate levels and in inter- psychosis stage. action with warmth can act as a protective factor and improved func- In a small exploratory study aimed at examining family burden in tioning over time. These authors also found that ARMS patients' the early stages of psychosis, Wong et al. (2008) indicated that levels perception of parental criticism predicted a worsening of attenuated of both subjective and objective burden were comparable between positive symptoms at follow-up. Moreover, interview-based ratings of ARMS and FEP caregivers. Family members also report that they help relatives' hostility and criticism significantly predicted a change in pos- patients with their everyday activities yet deny any resentment or itive symptoms at follow-up. In another study, O'Brien et al. (2008) subjective burden. They also denied the need to monitor or control showed that caregivers' warmth predicted an improvement in social behaviours in patients. Worry was a common feeling among family functioning while caregivers' positive remarks were associated with a members, but their lives have not been altered, and they do not have decrease in negative symptoms 4 months later. Besides, it was much anger or resentment. observed that parents who expressed more positive remarks regard- Two studies have focused on coping styles but from different per- ing their ill family member during the CFI tended to exhibit more spectives. On the one hand, Gerson et al. (2011) explored self-reported constructive behaviours during an interactional task. Besides, criticism coping styles in a small cohort of families of individuals with ARMS and was positively related to parents' conflict engaging behaviours during FEP. They found that families reported moderate use of ‘approach’ cop- the interactional task. Finally, the behaviours exhibited by relatives ing (such as planning, social support seeking, positive reinterpretation, during the interactional task were not predictive of ARMS patients' acceptance and turning to religion) and little use of ‘avoidant’ coping symptoms/functioning at a 4-month follow-up. However, in a strategies (such as denial/disengagement and alcohol and drug use). On 6-month follow-up report, O'Brien et al. (2009) showed that ARMS the other hand, Yee et al. (2020) focused on how ARMS individuals individuals' skilful problem-solving and constructive communication as cope with stress with their relatives as one of the most significant well as caregivers' constructive communication exhibited during face- stressors in the family environment. Specifically, this study examined to-face problem-solving discussions were associated with enhanced differences in coping strategies (i.e., engagement and disengagement) in social functioning of patients at follow-up. Conversely, ARMS individ- reaction to family stressors between ARMS individuals and healthy con- uals' conflicting communication during the problem-solving discussion trols, as well as their association with social support and clinical symp- with their parents was associated with increased positive symptoms toms. In comparison with the control group, ARMS individuals did not at follow-up. differ in their use of most engagement coping strategies (such as In their longitudinal study of 32 ARMS from a non-clinical problem-solving, emotion regulation, positive thinking, cognitive change population, Wang et al. (2015) found that self-reported family func- and distraction) but were more likely to use disengagement strategies tioning, specifically, better perceived problem-solving and affective (i.e., avoidance, denial and wishful thinking) when coping with family responsiveness from their parents predicted less severe positive and stress. Engagement strategies were linked to higher perceptions of negative symptoms at 6-month follow-up. Perceived family cohesion social support (in the form of having a source to go to for advice and and adaptability were negatively associated with general symptoms at guidance), whereas the use of disengagement strategies was associated baseline but were also negatively associated with general and with anxiety and depression symptoms, as well as with lower percep- disorganized symptoms at follow-up. tions of family support and increased family strain. As part of the European Prediction of Psychosis Study (EPOS) Other studies have focused on exploring aspects related to family designed to develop a prediction model of psychosis, Haidl functioning in the ARMS stage. Santesteban-Echarri et al. (2018)  ET AL. HINOJOSA-MARQUES 21 examined differences in family functioning, family satisfaction and com- 12 months after taking part in Izon et al. (2019). Ten of the munication in a sample of participants at risk of severe mental illness 14 caregivers completed the follow-up interview. Over 12 months, across different clinical stages (i.e., non-help-seeking youth with risk fac- four factors were important for caregivers to facilitate their caring tors, help-seeking youth with early mood and anxiety symptoms and role: (a) looking after their well-being; (b) accessing additional sup- distress, and youth with an attenuated psychiatric syndrome) and a con- port from family intervention; (c) communicating openly with the trol group. Findings showed that families in all groups fell within ranges individual; and (d) engaging with services for the individual. All these that reflected healthy family functioning. Moreover, all the groups aspects were important in improving family communication, meeting reported moderate levels of family communication, suggesting that par- caregivers' unmet needs and helping them feel more confident and ticipants generally feel good about communication patterns within the less isolated in their caring role. family. However, family satisfaction was lower in youth at risk for severe mental illness, who present with early signs of mood, anxiety or subthreshold psychotic symptoms, than other participants. Likewise, a 3.5 | Effect of family interventions on recent study (Salinger et al., 2018) sought to examine whether families symptomatic and/or functional ARMS outcomes of youth at clinical high risk for psychosis or bipolar disorder differ in communication patterns during problem-solving discussions. Results Several trials including family interventions for ARMS populations have indicated that mothers of youth at risk for psychosis displayed signifi- been developed, most of which have been conducted within the eight- cantly more conflictual and less constructive communication than those site North American Prodrome Longitudinal Study (NAPLS). The family- of youth at risk for bipolar disorder. focused therapy, consisting of an 18-session intervention that included The following studies used the qualitative methodology approach explicit communication and problem-solving training, was administered to explore relatives' experiences and family factors associated with to CHR (FFT-CHR) relatives and patients. Results from O'Brien caring for an ARMS family member. Baron et al. (2019) examined the et al. (2014) indicated that ARMS individuals and their respective fami- lived experience of associative stigma in 12 parents of adolescents at lies showed greater improvement from baseline to 6-month follow-up clinical high risk for psychosis. This study aimed to empower families in constructive communication (such as active listening) and greater to tell their stories and use their narrative to make recommendations decreases in conflicting behaviours (such as criticism) than those who and improve families' experiences of coping with clinical high risk for participated in a brief psychoeducational intervention. Importantly, the psychosis identification. Perceptions of stigma and coping with stigma FFT-CHR was found to be effective in reducing high-EE attitudes and emerged as the main domains according to parents' experiences. The improving positive communication within the family. Miklowitz narrative of the participants showed that (a) there is an increased et al. (2014) examined the effects of the FFT-CHR in reducing the risk experience of the stigma associated with psychotic-like symptoms as for psychosis onset among 129 ARMS individuals. Patients undertaking compared to other psychiatric symptoms; (b) parents attempt to bal- FFT-CHR showed greater improvement in positive symptoms from ance combatting stigma with the privacy needs of adolescents; and baseline to 6 months than those assigned to a brief family psycho- (c) stigma impacted treatment-seeking and participation in family educational intervention. Changes in psychosocial functioning over time groups that were seen as both potentially supportive and threatening. were age dependent, such that patients over 20 demonstrated greater These findings indicate that parents of ARMS children are probably functional improvement in FFT-CHR, whereas those between the ages reconciling their vulnerability to associative stigma with their notions of 16 and 19 showed greater functional improvement in the brief psy- of mental illness, which may include stigmatizing beliefs. The other choeducational intervention. O'Brien et al. (2015) explored whether two qualitative studies (Izon et al., 2019, 2020) were conducted FFT-CHR can reduce levels of perceived criticism and whether with the same sample of family/carers belonging to a larger Individ- decreases in perceived criticism predicted an improvement in symp- ual and Family Cognitive Behavioural Therapy trial study (Law toms. Findings indicated that perceived criticism reduced from baseline et al., 2021). Izon et al. (2019) explored the barriers and facilitators to 6 months for both treatment groups (FFT-CHR and brief family psy- for those supporting ARMS individuals through semi-structured choeducational intervention). A reduction in ARMS individuals' per- interviews with 14 caregivers. Findings suggest that caring for ceived criticism from baseline to 6 months predicted improvement in someone with ARMS can have an emotional and psychological attenuated positive symptoms at 12 months over and above symptom impact, as caregivers reported high levels of worry, uncertainty, dis- improvement at 6 months. tress, anger, feeling unsupported by health services, as well as In a pilot study with 337 youth (age 12–25) at risk of psychosis depression symptoms and suicidal feeling, as needing the most (McFarlane et al., 2015) from the multisite trial of the Early Detection immediate support from services. In summary, key barriers included and Intervention for the Prevention of Psychosis Program in the United relatives' unmet needs and limited confidence in accessing and pro- States, participants were assigned to different treatment groups based viding support, whereas facilitating factors included open communi- on the severity of positive symptoms. Individuals at the highest risk of cation with the individual, flexibility, understanding employers and psychosis transition (n = 250) were assigned to a Family-Aided Asser- feeling that they were supported. A further follow-up qualitative tive Community Treatment (FACT), while those with clinically lower risk study (Izon et al., 2020) was conducted to understand the different (n = 87) were assigned to community care. Although there were no dif- types of changes, similarities and reflections of those caregivers ferences in transition, family treatment had certain beneficial effects 22  ET AL. HINOJOSA-MARQUES over community care, as shown by the reduction of positive, negative, Hooley, 1998; Cechnicki et al., 2013; Marom et al., 2005; Wearden disorganized and general symptoms, increases in global functioning and et al., 2000), some of the reviewed cross-sectional studies aimed to greater overall improvement. Over the 24-month follow-up period, the explore the factors contributing to the early manifestation of EE. One FACT group increased their level of participation at work or school. line of enquiry has been the examination of whether ARMS' concur- Besides, effects for symptoms, functioning and global outcomes proved rent clinical status impacted on relatives' EE (Hamaie et al., 2016; to be larger for the group having very early psychosis than in the clinical McFarlane & Cook, 2007; Meneghelli et al., 2011) since it has been high-risk group (McFarlane et al., 2015). suggested that patients' clinical features may predict relatives' levels A further small (N = 6), open, uncontrolled trial pilot study (Landa of EE (for a review, see Hooley, 2007; Miklowitz, 2004). With excep- et al., 2016) examined a different Group- and Family-based Cognitive tion of McFarlane and Cook (2007), the rest of the reviewed studies Behavioural Therapy (GF-CBT) intervention intended to facilitate psy- showed that relatives' high-EE status did not seem to be reactive to chosocial recovery, decrease symptoms and prevent transition to ARMS patients' poor clinical status (Hamaie et al., 2016; Meneghelli psychosis in at-risk youth. Findings showed statistically significant et al., 2011). However, as suggested by Smith et al. (2018), it seems decreases in attenuated psychotic symptoms, negative symptoms, that other patient-related variables, not restricted to clinical character- depression and improvements in functioning in at-risk youth, whereas istics, such as patient-initiated violence, contribute to the expression family members showed significant improvements in the use of CBT of higher levels of criticism and hostility in relatives of ARMS patients. skills, enhanced communication with their offspring and greater confi- It is important to note that most of the cross-sectional studies dence in their ability to help. However, the results should be taken conducted in FEP samples also show inconsistent results regarding with caution because this study has important limitations including this issue. Although patients' poor clinical and functional status has lack of control group, unblinded assessments and potential confounds been related to an increase in relatives' EE in some FEP studies and the short 3-month follow-up. (King, 2000; Koutra et al., 2016; Mo et al., 2007), other studies have A recent study from a single-blind, pilot randomized controlled suggested that patients' symptoms/functioning have limited or no trial comparing a Combined Individual and Family Cognitive Behav-  impact on relatives' EE (Alvarez-Jiménez et al., 2010; Bachmann ioural Therapy intervention with treatment as usual in ARMS individ- et al., 2002; Heikkilä et al., 2002; Raune et al., 2004). The differences uals and their key caregiver reported the results of the rationale, in results among the above-mentioned studies (possibly due to the design and baseline characteristics of the study feasibility, but, given variability of the samples, measures or study design) leave unan- that it is currently underway, no data are yet available on the effec- swered the question as to what extent EE is a reaction to the severity tiveness of the trial (Law et al., 2021). of the relative's psychotic disorder (Heikkilä et al., 2002). As suggested by Hooley (2007) and Miklowitz (2004), understand- ing EE as a unilateral reaction on the part of a relative to a patient's clini- 4 | DISCUSSION cal characteristics is an incomplete approach to studying the ontogenesis of EE. In fact, developmental as well as bidirectional pro- To the best of our knowledge, this is the first review examining the wide cesses are involved in the genesis of high-EE attitudes (Hooley, 2007). range of factors related to family functioning in the ARMS stage. We There is therefore still an urgent need for prospective designs to deter- sought to identify empirical studies that have explored (1) relationships mine whether EE levels increase over time as a consequence of contin- of EE and other family environment constructs with ARMS symptoms/ ued exposure to ARMS patients' poor clinical and functional status. functioning and/or relatives' psychological variables; (2) the predictive Moreover, given the above-mentioned results from Smith et al. (2018), value of EE and other family environment constructs on ARMS symp- it would be necessary to examine a broader range of patients' variables toms and/or functional outcomes; (3) family environmental factors concerning EE, including their psychological and emotional state. How- (other than EE) explored in ARMS families; and (4) the effect of family ever, caution must be taken to avoid an oversimplistic view of EE. It is interventions on clinical and/or functional ARMS outcomes. therefore of the utmost importance to include relatives' psychological This section is intended to present a theoretical discussion of the variables in the early manifestation of EE. Following this line, Hinojosa- findings based on several questions addressed by the existing theoreti- Marqués et al. (2020) confirm that, as in Schizophrenia, relatives' emo- cal models of EE to analyse their usefulness in the understanding of the tional state and their cognitive representation of psychosis play an early emergence of EE at the risk stage for psychosis. For the remaining important role in the emergence, expression and maintenance of emo- reviewed literature that cannot be dealt with using this theoretical per- tional attitudes towards the patient, over and above patients' poor clini- spective, a more general discussion of findings will be provided. cal and functional status in the early stages of psychosis. 4.1 | Is EE a reaction to ARMS patients' clinical and 4.2 | Is EE related to ARMS relatives' psychological functional characteristics? variables? Given that high-EE environments have consistently been related to The scant evidence available suggests that a slightly different picture poor outcomes among patients with Schizophrenia (Butzlaff & emerges when EE is considered as a more relational variable in which  ET AL. HINOJOSA-MARQUES 23 relatives' cognitive representations of the disorder or generalized symptoms are within the patients' control (McNab et al., 2007; Vas- stress reactions contribute to its emergence. Unfortunately, the concelos e Sa et al., 2013). Nevertheless, the attributional model of EE sparse literature available on relatives' psychological factors associ- is a strictly cognitive-oriented paradigm that does not consider the piv- ated with EE precludes drawing meaningful conclusions. otal role of emotional and affective factors influencing the psychological Some of the reviewed studies based on the attributional model in experience of relatives in at-risk stages. It should therefore be borne in Schizophrenia (Barrowclough et al., 1994; Barrowclough & mind that relatives of ARMS patients are exposed, probably for the first Hooley, 2003) highlight the significant role of relatives' illness attribu- time, to early signs of psychosis. This potential threat may not only lead tions as mediators of the association between relatives' EE and early to creating cognitive appraisals about the causes of the disorder but also psychosis patients' clinical/functional features (Domínguez-Martínez significant affective reactions to a recently processed situation. et al., 2014) or as cross-sectional predictors of criticism (Domínguez- According to Hinojosa-Marqués et al. (2020), both cognitive and emo- Martínez et al., 2017). To expand these preliminary cross-sectional tional factors, and even their interaction, are important predictors of the reports by using an extended sample and a longitudinal design, emergence and maintenance of EE in the ARMS period. Hinojosa-Marqués et al. (2020) tested a more comprehensive predic- To date, relatives' emotional states (such as distressing affective tive model of EE in early psychosis (ARMS and FEP individuals) by states) and its relationship with EE have rarely been studied in the including both patients' illness-related variables and relatives' psycho- pre-psychotic phase, which precludes the possibility of examining logical factors. Findings confirmed that relatives' psychological dis- whether EE arises from the perceived stress related to the caregiving tress and negative illness attributions predicted EE dimensions (over role, as suggested by the ‘carer appraisal model of EE’ based on FEP and above patients' clinical and functional features) across time at samples (Kuipers et al., 2006; Raune et al., 2004). Some of the studies both subclinical and onset stages of psychosis, lending further support that have examined these relationships presented contradictory find- to the attributional model (Barrowclough & Hooley, 2003). This result ings. Hamaie et al. (2016) only observed links between relatives' also indicates that the emergence of EE attitudes also stems from neg- depression and criticism in FEP relatives but not in ARMS relatives. ative emotional responses and negative ‘hot’ cognitions in the ARMS However, similar proportions of mild to moderate depressive symp- and recent-onset stages of psychosis. toms were reported for both ARMS and FEP relatives. Authors Hinojosa-Marqués, Domínguez-Martínez, bal-Narváez, Cristo suggested that the interaction between caregivers' emotional distress et al. (2019) also found a similar pattern of associations between EE and criticism may emerge after the onset of psychosis but not at the dimensions (assessed by both momentary and psychometric self-reports at-risk stage. Conversely, Domínguez-Martínez et al. (2017) found dif- [FQ]) and relatives' illness attributions in daily life. Some of these associ- ferences between the ARMS and FEP relatives regarding the relation- ations are in line with previous early psychosis studies (e.g., Bolton ship between EE indices and anxiety but not with depression. et al., 2003; McNab et al., 2007; Vasconcelos e Sa et al., 2013) and sup- Moreover, Hinojosa-Marqués et al. (2020) found strong associations port the attributional model of EE (Barrowclough & Hooley, 2003). between relatives' baseline levels of anxiety and depression and rela- Hinojosa-Marqués, Domínguez-Martínez, bal-Narváez, Cristo tives' EE dimensions at both baseline and 6-month follow-up. These et al. (2019) argued that at the early stages of psychosis, many relatives findings contradict the assumption by Hamaie et al. (2016) by showing may still exhibit low-defined illness attributions due to confusion, uncer- that relatives' emotional states are strongly linked to EE attitudes at tainty and the lack of knowledge about the disorder. Thus, relatives the at-risk and onset stages of psychosis, even at 6-month follow-up. may believe they can control the disorder by themselves and, at the Indeed, relatives' levels of anxiety and depression proved to be predic- same time, that patients can have significant control over their behav- tors of critical attitudes in both ARMS and FEP groups. The difference iour. As the disorder progresses, relatives may express more clearly in results between these studies can be explained by cultural differ- defined illness attributions which, in turn, would result in more specific ences between the Japanese (Hamaie et al., 2016) and Spanish sam- behavioural reactions. Presumably, the high emotional impact of early ples (Domínguez-Martínez et al., 2017; Hinojosa-Marqués psychosis on family members may lead them to a low understanding of et al., 2020). However, more studies are required on the association their feelings (i.e., diminished emotional clarity). Low levels of emotional between EE and relatives' levels of anxiety and depression at the early clarity may influence the way they interpret patients' behaviours, thus stages of psychosis to be able to theorize about it. causing relatives' cognitive ambivalence regarding the control of the dis- Following this line, Hinojosa-Marqués, Domínguez-Martínez, bal-Narváez, order (Hinojosa-Marqués, Domínguez-Martínez, Cristo  bal-Narváez, et al. (2019) showed that relatives' EE attitudes are Cristo et al., 2019). According to Domínguez-Martínez et al. (2017), the attri- closely related to momentary negative affective experiences, increased butional model of EE should be tailored to the developmental specific- appraisals of situational stress and feeling burdened by the patient and ities of early psychosis thereby including the differences related to the decreased positive affect in daily life at both the subclinical and onset various stages which families undergo across the psychosis continuum. stages of the illness. Overall, these results seem to be partially consis- This would require integrating the critical role of emotional factors tent with previous early psychosis findings indicating an association influencing the psychological experience of relatives in at-risk stages. between EE and relatives' distress and/or burden (Domínguez-Martínez To date, the attributional model of EE has received some empirical et al., 2017; Hinojosa-Marqués et al., 2020; Raune et al., 2004; support at the recent-onset stage of psychosis. For example, families of Tomlinson et al., 2014) but do not replicate previous early psychosis FEP patients have shown that critical relatives tend to believe that research suggesting that EOI is more closely related to distress and 24  ET AL. HINOJOSA-MARQUES  burden than criticism (Alvarez-Jiménez et al., 2010; González-Blanch within the family (O'Brien et al., 2009) and positive self-reported family et al., 2010; Jansen, Gleeson, & Cotton, 2015). functioning (Wang et al., 2015) were predictors of improvement in Unfortunately, the study of relatives' distress and EE in the high- ARMS patients' symptoms and/or functioning over time. As suggested risk period is still very scarce compared with FEP literature. Converg- by a recent review (Butler et al., 2019), the protective effects of positive ing evidence suggests that greater levels of psychological distress in family attitudes on symptomatic relapse are most evident at the ARMS  FEP caregivers appear to be related to increased levels of EE (Alvarez- stage. Indeed, it was argued that the reduced chronicity of co-morbid Jiménez et al., 2010; Jansen, Gleeson, & Cotton, 2015; Jansen, Haahr, difficulties (such as social anxiety, low mood and substance use; et al., 2015; Koutra et al., 2014; Raune et al., 2004; Tomlinson McGorry & Yung, 2003) could imply that there is greater potential for et al., 2014). There is therefore an urgent need for more research the protective effects of positive family environments in the early aimed at examining caregivers' distress during the pre-psychotic stage course of psychosis (Butler et al., 2019). to achieve a better understanding of the impact of prodromal phases Another differential aspect of the ARMS stage is that relatives' EOI of illness on relatives' mental well-being. appeared to act more as a protective factor, given that it was found to Only one study (Hinojosa-Marqués, Domínguez-Martínez, be related to improvement in patients' clinical/functional status over time Sheinbaum, et al., 2019) pointed to the role of relatives' perceived loss (O'Brien et al., 2006; Schlosser et al., 2010). In this line, it is important to as a major driver of EE attitudes in a mixed sample of ARMS and FEP assess the differential predictive power of the two EE components (criti- relatives. Moreover, results from this study expanded previous find- cism and EOI) since criticism has been found to make a greater contribu- ings from the FEP literature (Patterson, 2013; Patterson et al., 2000, tion to symptomatic relapses in Schizophrenia (Cechnicki et al., 2013; 2005), showing that relatives' attachment anxiety may be a mediating Kavanagh, 1992; Marom et al., 2005) and even in the FEP period (Koutra mechanism whereby perceived loss influences the manifestation of et al., 2015). However, results regarding the predictive value of criticism both criticism and EOI attitudes in at-risk stages. This study consti- in ARMS samples are somewhat mixed: while one study found that rela- tutes a novel contribution, but further research focused on the early tives' criticism predicted a worsening of attenuated psychotic symptoms grief reactions of ARMS family members is required. over time (Schlosser et al., 2010), others did not report the same associa- tion (Haidl et al., 2018; O'Brien et al., 2006). Given the controversial rela- tionship between criticism and patients' clinical attributes in the high-risk 4.3 | Are EE or other family-related variables period, further study is required to determine whether cumulative expo- influencing clinical and/or functional ARMS sure to critical attitudes can impact negatively on ARMS symptoms characteristics? and/or functioning. Furthermore, Haidl et al. (2018), the first study exploring the predictive value of perceived EE in conversion into FEP, Beyond studies focused on disentangling the patients and/or relatives' highlighted how other valuable constructs, such as patients' perceived factors involved in the early development of EE, another parallel line irritability of a key relative, were predictive of ARMS conversion into of research has focused on testing the impact of EE on ARMS FEP. These significant results should encourage future studies to adopt a patients' concurrent symptoms and/or functioning (Carol & more holistic approach to studying family environment variables contrib- Mittal, 2015; Domínguez-Martínez et al., 2014; Tsai et al., 2015). uting to relapse, rather than focusing solely on relatives' criticism and Although cross-sectional studies reported mixed results con- EOI attitudes, including positive affect that showed to be protective of cerning this issue, it seems that relatives' criticism is more closely relapse in FEP individuals (Lee et al., 2014). associated with patients' concurrent clinical and functional status than From a diathesis-stress model perspective, stress within the fam- EOI. However, the scant research available makes it difficult to draw ily environment has been shown to contribute to the development firm conclusions. Overall findings on the cross-sectional associations and maintenance of symptoms in psychotic disorders (Hooley & between EE and family functioning (as rated by patients or relatives) Gotlib, 2000). Recently, research interest in family stress, coping strat- with patients' status are inconclusive in the ARMS literature. How- egies and the parent–child relationship at the ARMS stage has ever, it should be noted that not all studies on the family environment increased. It has been demonstrated that families reported moderate in Schizophrenia and/or FEP samples find correlations between levels use of ‘approach’ coping and occasional use of ‘avoidant’ coping  of EE and the severity of patients' concurrent clinical status (Alvarez- strategies (Gerson et al., 2011), whereas ARMS individuals were more Jiménez et al., 2010; Bachmann et al., 2002; Cutting et al., 2006; likely to use disengagement strategies (such as avoidance, denial and Heikkilä et al., 2002; Vasconcelos e Sa et al., 2016). Accurate predic- wishful thinking) to cope with family stress that is associated with less tions of the influence of EE on ARMS patients' status are probably social support and greater anxiety and depression, potentially com- impossible to identify using cross-sectional designs. pounding the social stressors and clinical burden these individuals are Regarding the reviewed studies exploring the longitudinal impact of experiencing (Yee et al., 2020). family environment variables on ARMS symptomatic relapse, it appears Moreover, from a parent–child relationship perspective, Peh that most of the research reviewed highlighted the role of positive fam- et al. (2020) provide empirical evidence supporting the hypothesis ily aspects. Interestingly, studies revealed that (1) relatives' positive EE that parent–child relationships are associated with the ARMS state. components (such as positive remarks and warmth) (O'Brien Specifically, they suggest that an affectionless-overprotective- et al., 2006, 2008) and (2) observed positive interactional patterns controlling parenting style could not be beneficial for young people  ET AL. HINOJOSA-MARQUES 25 who are vulnerable to psychosis. Using similar measures, further literature (Gerson et al., 2011; Wong et al., 2008), it would appear that cross-sectional studies showed that other family variables (such as relatives of ARMS patients are not yet as negatively affected by the nurturing behaviour, problem-solving and quality of parent–child rela- disorder as they are in chronic stages, although they already show tionships), as rated by patients' perceptions, were unrelated to psy- the emotional and psychological impact of their caring role (such as chotic symptom severity in both ARMS and FEP groups (Welsh & higher levels of worry, fear, distress, anger, anxiety and depressive Tiffin, 2015), except for the association between the EE subscale and symptoms) that are closely linked to the lack of understanding, uncer- patients' manic symptomatology. However, a different picture tainty, feelings of lack of support from health services and unmet emerges when considering ARMS patients' family perceptions as mod- needs (Izon et al., 2019). erators. Bentley et al. (2016) found that in the ARMS group, the sever- The only study focused on examining family functioning from the ity of social stress is dependent on the patient's perceived quality of patients' perspective (Santesteban-Echarri et al., 2018) indicated that the parent–child relationship. Likewise, Thompson et al. (2019) ARMS patients reported lower family satisfaction compared with showed the moderating effect of ARMS patients' perceived family healthy controls or non-help-seeking participants with risk factors for functioning between symptoms and social/role functioning. mental illness. However, ARMS patients reported healthy family func- tioning, in terms of cohesion and adaptability within the family. In contrast to these findings, another of the reviewed studies indicated 4.4 | To what extent is the family emotional that ARMS individuals scored less than the control group on perceived climate affected by the ARMS stage? cohesion and adaptability within the family (Wang et al., 2015) and reported poorer perceived problem-solving and lower levels of nurtur- Studies exploring family factors as the main outcome are still very ing behaviour in their families compared to controls (Welsh & scarce in the ARMS literature but have been increasing in recent Tiffin, 2015). Moreover, Salinger et al. (2018) indicated that communi- years. Only eight studies focused on examining family-related vari- cation patterns also appeared to be affected. ables regardless of relatives' EE levels or patients' symptom severity Although more research is required to determine the generaliz- (Baron et al., 2019; Gerson et al., 2011; Izon et al., 2019, 2020; Salin- ability of the findings, dimensional approaches, which adopt a broader ger et al., 2018; Santesteban-Echarri et al., 2018; Wong et al., 2008; view of family functioning in terms of cohesion and adaptability and Yee et al., 2020). Wong et al. (2008), the first study examining burden other global indicators of family welfare (such as problem-solving, in the ARMS stage, showed that family burden was comparable communications patterns and nurturing behaviour), have been between ARMS and FEP families. ARMS families were characterized suggested as being more useful and comprehensive for capturing the by worry and active involvement in care, but their lives had not yet potential disruptions of the family environment caused by the disor- been disrupted by the disorder, which only had a limited impact on der (Koutra et al., 2015). their daily routines. One possible explanation for these results could Finally, findings from Baron et al. (2019) expand the literature on be that in the ARMS stage, characterized by less symptom severity the topic of caregivers of adolescents at risk for psychosis and provide and chronicity, relatives' active involvement or worry is not yet associ- a different perspective through the narrative of parents on their per- ated with significant emotional disturbances in their lives. Instead, ceptions of stigma and how they cope with it. This is the first qualita- findings from the FEP literature suggest that high levels of burden and tive study to focus on this topic in the at-risk stage of psychosis. The psychological distress are already present after the recent onset of authors pointed out that the ‘psychosis risk’ label carries more stigma psychosis (Addington et al., 2003; Boydell et al., 2014). Continued than other more common mental health disorders, which may create a exposure to psychopathology or poor functioning, as well as the care- barrier to seeking appropriate treatment. givers' emotional shock caused by the early onset of the illness, there- fore probably lead to the reinforcement of burden and/or distress responses. 4.5 | Do family-based interventions improve Findings from Gerson et al. (2011) also supported the fact that outcomes in the ARMS stage? ARMS relatives are characterized by active involvement in care, as they reported moderate use of ‘approach’ coping (such as planning, In comparison with longer term illness (Barrowclough & Tarrier, 1990; seeking social support, positive reinterpretation and acceptance) and Berglund et al., 2003; Chien & Norman, 2009) and recent-onset psy- only occasional use of ‘avoidant’ coping strategies (such as denial/dis- chosis groups (for a review, see Askey et al., 2007; Bird et al., 2010; engagement and alcohol and drug use). Avoidant coping strategies Claxton et al., 2017; Penn et al., 2005), there have been fewer family- have been more frequently described in relatives of patients with based intervention studies at the ARMS stage and the evidence based chronic Schizophrenia (Fortune et al., 2005; Huang et al., 2008), which on its efficacy during this pre-illness stage is still limited. Some early may indicate that these types of coping strategies become more prev- intervention programmes include multifamily psychoeducation, indi- alent over time as a consequence of relatives' increased fatigue or vidual and/or group family interventions as part of the integrated burden. treatment focused on preventing progression to FEP. However, some To date, little is known about relatives' burden and/or coping of them have reported the overall effects of the set of interventions reactions in the high-risk period. As suggested by the previous on outcomes, which limits the interpretation of the specific effect of 26  ET AL. HINOJOSA-MARQUES family intervention in preventing the transition to psychosis therefore still an enormous need for research aimed at exploring the (e.g., Bechdolf et al., 2012; Nordentoft et al., 2006). The reviewed needs of caregivers regardless of their EE attitudes or the course of studies suggest that, as in Schizophrenia or FEP stages, family-based illness of their affected relative. interventions are effective in reducing both relatives' high-EE atti- tudes (O'Brien et al., 2014) and patients' perceived levels of criticism (O'Brien et al., 2015) as well as in improving family communication 4.7 | Clinical implications and ARMS patients' clinical and functional outcomes over time (Landa et al., 2016; McFarlane et al., 2015; Miklowitz et al., 2014; O'Brien Although a great deal of research on family-based interventions with et al., 2015). The primary objective of the reviewed family-based ARMS populations remains to be done, the evidence summarized in interventions at the ARMS stage was to delay or prevent the this review concerning the family treatment in ARMS stages suggests transition to psychosis by reducing the negative aspects of that early interventions benefit individuals at risk for psychosis as well EE. However, given the potentially protective effects of positive as family dynamics. Family work with ARMS relatives is therefore family attitudes (such as warmth, positive remarks and positive inter- essential for maximizing the adaptive functioning of the family and actional patterns; O'Brien et al., 2006, 2008, 2009; Wang et al., 2015), minimizing disruption to family life and the risk of ARMS individuals' it would also appear relevant to implement treatment strategies to deterioration given their high vulnerability to environmental stressors foster the positive aspects of the family environment (Butler (Fusar-Poli et al., 2017). et al., 2019). Findings of this review showed that relatives' own needs and the emotional impact of caregiving are still a neglected intervention area in the ARMS stage. An increased focus on the impact of the at-risk 4.6 | Methodological limitations and future psychosis stage on relatives' mental well-being is required to expand research family support based on their psychological needs and to clarify the mechanisms leading to dysfunctional family dynamics during the criti- We identified several important methodological limitations in the cal ARMS period, considering other family factors beyond EE. Some reviewed literature. First, there was a high degree of heterogeneity of the identified barriers for family members include a lack of under- among the studies included. There were differences regarding study standing of individuals' symptoms, limited confidence in their ability designs, methods of assessment (family and patients' measures), dif- to support them, unproductive coping strategies impacting their ferent types of mixed samples, follow-up or other features that make health and/or well-being (Izon et al., 2019) and the associated stigma the comparison between studies difficult due to the heterogeneity to ‘psychosis risk’ label, which may create a barrier to seeking appro- between them. Given the characteristics of the present theoretical priate treatment (Baron et al., 2019). Besides, it has been widely dem- and comprehensive review, we do not provide an analysis of the onstrated that psychosis impacts family members' emotional state and methodological quality of the studies, which would be important to family dynamic throughout the different stages of the disorder. Con- address in further systematic reviews using quality appraisal tools. sequently, it is necessary to develop family resources and alternative Second, most of the studies were cross-sectional with small sam- intervention approaches, to educate, normalize and validate their own ple sizes. Future studies could shed light on variations in patient– experiences. Izon et al. (2020) suggest some aspects to consider in family dynamics over time by using longitudinal designs and large clinical settings to facilitate relatives' role as caregivers, improve family samples of ARMS patients and their respective relatives. Third, most communication and improve their quality of life: (a) exploring care- of the reviewed studies recruited combined samples of ARMS and givers' coping strategies to manage their worries, anxiety and distress FEP relatives. Although some of them reported separate results for can help them prioritize their health and well-being; the ARMS sample or offered comparative results between the sam- (b) psychoeducation and normalizing caregivers' experiences can help ples, this precluded a more detailed examination of the different them feel validated and reduce their worry; (c) facilitating open com- aspects of ARMS relatives concerning family environment variables. munication between the patient and their family can positively impact Another limitation is that EE and/or other family environment con- on their relationship; (d) early engagement between services and structs were assessed based on either relatives' or patients' perspec- ARMS individuals can help families feel less isolated; and (e) when fea- tives. Only two studies assessed both patients' and relatives' sible, services should invite/involve all the family members of ARMS perceptions of EE (O'Brien et al., 2015; Schlosser et al., 2010). Given individuals in sessions and explore family support strategies in manag- that EE reflects a transactional process between patients and relatives ing their distress. Furthermore, Baron et al. (2019) also suggest that (Strachan et al., 1989), further studies should consider the importance stigma should be included as an important topic to work on within of obtaining information from both relatives' and patients' perspec- psychoeducation and family therapeutic groups to help families cope tives. A final limitation is that only a few studies considered the role with it and prevent a delay in adequate patient treatment. Besides, of relatives' psychological features (such as illness beliefs, distress and alternative interventions approach, such as online programmes and stress reactions) in explaining EE attitudes. Unfortunately, this pre- forums that families can access conveniently, asynchronously cluded a detailed discussion of the usefulness of existing theoretical and more anonymously, may be useful in the beginning to prevent models for explaining EE in the at-risk stage of psychosis. There is associative stigma in parents.  ET AL. HINOJOSA-MARQUES 27 Furthermore, it has been highlighted the importance of offering longitudinal studies regarding the contribution of criticism to ARMS caregivers about psychotic symptoms and tailors psychological inter- patients' clinical course. Moreover, EOI appears to act more as a pro- ventions according to the stage of the disorder (considering that psy- tective factor and could have some positive effects on patients' out- chotic disorder is not necessarily the final outcome). This would help comes at the ARMS stage. Finally, family-based interventions have relatives to handle difficult thoughts, negative appraisals, and distress, demonstrated enormous potential for reducing high-EE attitudes to adequately cope and address the challenges of the disorder over and/or improving ARMS patients' outcomes. Determining relatives' time (Domínguez-Martínez et al., 2017), especially when they are needs and supporting family caregivers also warrant further attention. dealing with the emergence of high-risk symptoms, where difficulties related to the caring role different from those with onset or well- AC KNOW LEDG EME NT S established psychotic disorder. This work is supported by the Spanish Ministerio de Economía y Competitividad (PSI2017-87512-C2-01) and the Comissionat per a Universitats i Recerca of Generalitat de Catalunya (2017SGR1612). 5 | C O N CL U S I O N S  Catalana de Recerca N. Barrantes-Vidal is supported by the Institucio i Estudis Avançats (ICREA) Academia Award and the Centro de Inves- Although much research is needed to gain a better understanding of n Biomédica en Red de Salud Mental (CIBERSAM), Instituto de tigacio family dynamics in the pre-psychotic stage, it is important to appreci- Salud Carlos III, Barcelona, Spain. This paper is based on work ate existing research efforts to encourage the study of the family risk appearing in a doctoral dissertation (Hinojosa-Marqués, 2018). factors across the psychosis continuum. The recent interest in this area is beneficial in the shift to a paradigm in family environment AUTHOR CONTRIBU TIONS research that will hopefully continue to attract the attention of All the authors listed have made a substantial, direct and intellectual researchers in the near future. contribution to the work and approved it for publication. In conclusion, the preliminary evidence summarized in this review reflects the state of the art of the study of family environmental fac- CONFLIC T OF INT ER E STS tors at the ARMS stage. Given the enormous heterogeneity of the The authors declare that the research was conducted in the absence studies reviewed, it is difficult to compare them all or to draw mean- of any commercial or financial relationships that could be construed ingful conclusions. The scant literature researching potential contrib- as a potential conflict of interest. uting factors to the early manifestation of EE has precluded the identification of the specific mechanisms underlying its ontogenesis. DATA AVAILABILITY STAT EMEN T Overall, relatives' high-EE status did not appear to be reactive to Data sharing does not apply to this article as no new data were cre- ARMS patients' poor clinical status. However, relatives' cognitive rep- ated or analysed in this study. resentations of psychosis may have an important role to play, either as mediators of the relationship between EE and ARMS clinical/func- OR CID tional features or as predictors of EE attitudes. However, there is still Lídia Hinojosa-Marqués https://orcid.org/0000-0002-1778-8522 a dearth of literature on relatives' psychological appraisals in terms of Tecelli Domínguez-Martínez https://orcid.org/0000-0003-4369- their relationship to EE. Likewise, relatives' distress and/or burden 8288 and its association with EE have been largely unexplored or even as Neus Barrantes-Vidal https://orcid.org/0000-0002-8671-1238 the main variables of interest at the ARMS stage, meaning that rela- tives' psychological needs in the prodromal period have been over- RE FE RE NCE S looked. Furthermore, the available literature has failed to achieve Addington, J., Coldham, E. L., Jones, B., Ko, T., & Addington, D. (2003). The consistent results when analysing the impact of EE on ARMS patients' first episode of psychosis: The experience of relatives. Acta Psy- concurrent symptoms and/or functioning. In the same vein, further chiatrica Scandinavica, 108(4), 285–289. https://doi.org/10.1034/j. 1600-0447.2003.00153.x dimensional family functioning indicators (such as cohesion, adaptabil-  Alvarez-Jiménez, M., Gleeson, J. F., Cotton, S. M., Wade, D., Crisp, K., ity, nurturing behaviour, problem-solving, quality of parent–child rela- Yap, M. B. H., & McGorry, P. D. (2010). Differential predictors of tionships and communication patterns) do not appear to impact critical comments and emotional over-involvement in first-episode negatively on psychotic symptom severity. However, some initial evi- psychosis. Psychological Medicine, 40(1), 63–72. https://doi.org/10. 1017/S0033291708004765 dence suggests the moderating role of these family functioning indica- Askey, R., Gamble, C., & Gray, R. (2007). Family work in first-onset psycho- tors in shaping the expression of ARMS patients' symptoms. Although sis: A literature review. Journal of Psychiatric and Mental Health Nursing, evidence is still limited, several longitudinal studies highlight the sig- 14(4), 356–365. https://doi.org/10.1111/j.1365-2850.2007.01093.x nificant role of positive family aspects as predictors of ARMS patients' Bachmann, S., Bottmer, C., Jacob, S., Kronmüller, K. T., Backenstrass, M., Mundt, C., Renneberg, B., Fiedler, P., & Schröder, J. (2002). Expressed clinical and/or functional improvement over time. Nevertheless, there emotion in relatives of first-episode and chronic patients with schizo- is an urgent need for studies aimed at determining the predictive phrenia and major depressive disorder—A comparison. Psychiatry value of EE dimensions (such as criticism and EOI) on the clinical out- Research, 112(3), 239–250. https://doi.org/10.1016/S0165-1781(02) come of ARMS patients. There is insufficient consistency across 00226-3 28  ET AL. HINOJOSA-MARQUES Baron, J., Salvador, M., & Loewy, R. (2019). Experience of Butzlaff, R., & Hooley, J. (1998). Expressed emotion and psychiatric associative stigma in parents of adolescents at risk for psychosis. Early relapse: A meta-analysis. Archives of General Psychiatry, 55, 547–552. Intervention in Psychiatry, 13(4), 761–766. https://doi.org/10.1111/ https://doi.org/10.1001/archpsyc.55.6.547 eip.12555 Caqueo-Urízar, A. C., Castillo, C. M., Giráldez, S. L., Maturana, S. L. L., Barrowclough, C., & Hooley, J. M. (2003). Attributions and expressed emo- Pérez, M. R., & Tapia, F. M. (2014). An updated review on burden on tion: A review. Clinical Psychology Review, 23(6), 849–880. https://doi. caregivers of schizophrenia patients. Psicothema, 26(2), 235–243. org/10.1016/S0272-7358(03)00075-8 https://doi.org/10.7334/psicothema2013.86 Barrowclough, C., Johnston, M., & Tarrier, N. (1994). Attributions, Carol, E. E., & Mittal, V. A. (2015). Resting cortisol level, self-concept, and expressed emotion, and patient relapse: An attributional model of putative familial environment in adolescents at ultra high-risk for psy- relatives' response to schizophrenic illness. Behavior Therapy, 25(1), chotic disorders. Psychoneuroendocrinology, 57, 26–36. https://doi. 67–88. https://doi.org/10.1016/S0005-7894(05)80146-7 org/10.1016/j.psyneuen.2015.03.018 Barrowclough, C., & Tarrier, N. (1990). Social functioning in schizophrenic Cechnicki, A., Bielanska, A., Hanuszkiewicz, I., & Daren, A. (2013). The pre- patients. I. The effects of expressed emotion and family intervention. dictive validity of Expressed Emotions (EE) in schizophrenia. A 20-year Social Psychiatry and Psychiatric Epidemiology, 25, 125–129. https:// prospective study. Journal of Psychiatric Research, 47(2), 208–214. doi.org/10.1007/BF00782739 https://doi.org/10.1016/j.jpsychires.2012.10.004 Barrowclough, C., Tarrier, N., & Johnston, M. (1996). Distress, expressed Chien, W. T., & Norman, I. (2009). The effectiveness and active ingredients emotion, and attributions in relatives of schizophrenia patients. Schizo- of mutual support groups for family caregivers of people with psy- phrenia Bulletin, 22(4), 691–701. https://doi.org/10.1093/schbul/22. chotic disorders: A literature review. International Journal of Nursing 4.691 Studies, 46(12), 1604–1623. https://doi.org/10.1016/j.ijnurstu.2009. Bechdolf, A., Wagner, M., Ruhrmann, S., Harrigan, S., Putzfeld, V., 04.003 Pukrop, R., Brockhaus-Dumke, A., Berning, J., Janssen, B., Decker, P., Claxton, M., Onwumere, J., & Fornells-Ambrojo, M. (2017). Do family Bottlender, R., Maurer, K., Möller, H. J., Gaebel, W., Häfner, H., interventions improve outcomes in early psychosis? A systematic Maier, W., & Klosterkötter, J. (2012). Preventing progression to first- review and meta-analysis. Frontiers in Psychology, 8, 371. https://doi. episode psychosis in early initial prodromal states. The British Journal org/10.3389/fpsyg.2017.00371 of Psychiatry, 200(1), 22–29. https://doi.org/10.1192/bjp.bp.109. Cole, J. D., & Kazarian, S. S. (1988). The level of expressed emotion scale: 066357 A new measure of expressed emotion. Journal of Clinical Psychology, Bentley, E., Millman, Z. B., Thompson, E., Demro, C., Kline, E., Pitts, S. C., 44(3), 392–397. https://doi.org/10.1002/1097-4679(198805)44:3% DeVylder, J. E., Smith, M. E., Reeves, G., & Schiffman, J. (2016). High- 3C392::AID-JCLP2270440313%3E3.0.CO;2-3 risk diagnosis, social stress, and parent-child relationships: A modera- Cutting, L. P., Aakre, J. M., & Docherty, N. M. (2006). Schizophrenic tion model. Schizophrenia Research, 174(1–3), 65–70. https://doi.org/ patients' perceptions of stress, expressed emotion, and sensitivity to 10.1016/j.schres.2016.04.014 criticism. Schizophrenia Bulletin, 32(4), 743–750. https://doi.org/10. Bentsen, H., Notland, T. H., Munkvold, O., Boye, B., Ultstein, I., Biørge, H., 1093/schbul/sbl001 Uren, G., Lerbryggen, A. B., Oskarsson, K. H., Berg-Larsen, R., Domínguez-Martínez, T., Medina-Pradas, C., Kwapil, T. R., & Barrantes- Lingjærde, O., & Malt, U. F. (1998). Guilt proneness and expressed Vidal, N. (2014). Relatives' illness attributions mediate the association emotion in relatives of patients with schizophrenia or related psycho- of expressed emotion with early psychosis symptoms and functioning. sis. British Journal of Medical Psychology, 71, 125–138. https://doi.org/ Psychiatry Research, 218(1–2), 48–53. https://doi.org/10.1016/j. 10.1111/j.2044-8341.1998.tb01374.x psychres.2014.04.012 Berglund, N., Vahlne, J. O., & Edman, Å. (2003). Family intervention in Domínguez-Martínez, T., Medina-Pradas, C., Kwapil, T. R., & Barrantes- schizophrenia—Impact on family burden and attitude. Social Psychiatry Vidal, N. (2017). Relatives' expressed emotion, distress and attribu- and Psychiatric Epidemiology, 38(3), 116–121. https://doi.org/10. tions in clinical high-risk and recent onset of psychosis. Psychiatry 1007/s00127-003-0615-6 Research, 247, 323–329. https://doi.org/10.1016/j.psychres.2016. Bird, V., Premkumar, P., Kendall, T., Whittington, C., Mitchell, J., & 11.048 Kuipers, E. (2010). Early intervention services, cognitive-behavioral Fortune, G., Smith, J. V., & Garvey, K. (2005). Perceptions of psychosis, therapy and family intervention in early psychosis: Systematic review. coping, appraisals, and psychological distress in the relatives of British Journal of Psychiatry, 197(5), 350–356. https://doi.org/10. patients with schizophrenia: An exploration using self-regulation the- 1192/bjp.bp.109.074526 ory. British Journal of Clinical Psychology, 44, 319–331. https://doi.org/ Bolton, C., Calam, R., Barrowclough, C., Peters, S., Roberts, J., 10.1348/014466505X29198 Wearden, A., & Morris, J. (2003). Expressed emotion, attributions and Friedmann, M. S., McDermut, W. H., Solomon, D. A., Ryan, C. E., depression in mothers of children with problem behavior. Journal of Keitner, G. I., & Miller, I. W. (1997). Family functioning and mental ill- Child Psychology and Psychiatry, 44, 242–254. https://doi.org/10. ness: A comparison of psychiatric and nonclinical families. New York, 1111/1469-7610.00117 36, 357–367. https://doi.org/10.1111/j.1545-5300.1997.00357.x Boydell, J., Onwumere, J., Dutta, R., Bhavsar, V., Hill, N., Morgan, C., Fusar-Poli, P., Tantardini, M., De Simone, S., Ramella-Cravaro, V., Dazzan, P., Morgan, K., Pararajan, M., Kuipers, E., Jones, P., Oliver, D., Kingdon, J., Kotlicka-Antczak, M., Valmaggia, L., Lee, J., Murray, R., & Fearon, P. (2014). Caregiving in first-episode psychosis: Millan, M. J., Galderisi, S., Balottin, U., Ricca, V., & McGuire, P. (2017). Social characteristics associated with perceived ‘burden’ and associa- Deconstructing vulnerability for psychosis: Meta-analysis of environ- tions with compulsory treatment. Early Intervention in Psychiatry, 8(2), mental risk factors for psychosis in subjects at ultra high-risk. European 122–129. https://doi.org/10.1111/eip.12041 Psychiatry, 40, 65–75. https://doi.org/10.1016/j.eurpsy.2016.09.003 Brown, G. W., Birley, J. L. T., & Wing, J. K. (1972). Influence of family Gerson, R., Wong, C., Davidson, L., Malaspina, D., McGlashan, T., & life on the course of schizophrenic disorders: A replication. British Corcoran, C. (2011). Self-reported coping strategies in families of Journal of Psychiatry, 121, 241–258. https://doi.org/10.1192/bjp.121. patients in early stages of psychotic disorder: An exploratory study. 3.241 Early Intervention in Psychiatry, 5(1), 76–80. https://doi.org/10.1111/j. Butler, R., Berry, K., Varese, F., & Bucci, S. (2019). Are family warmth and 1751-7893.2010.00251.x positive remarks related to outcomes in psychosis? A systematic González-Blanch, C., Martín-Muñoz, V., Pardo-García, G., Martínez- review. Psychological Medicine, 49, 1250–1265. https://doi.org/10.  García, O., Alvarez-Jiménez, M., Rodríguez-Sánchez, J. M., Vázquez- 1017/S0033291718003768 Barquero, J. L., & Crespo-Facorro, B. (2010). Effects of family  ET AL. HINOJOSA-MARQUES 29 psychoeducation on expressed emotion and burden of care in first- review. Psychiatry Research, 270(June), 661–672. https://doi.org/10. episode psychosis: a prospective observational study. The Spanish jour- 1016/j.psychres.2018.10.065 nal of psychology, 13(1), 389–395. https://doi.org/10.1017/ Izon, E., Berry, K., Law, H., Shiers, D., & French, P. (2020). “I don't think I S1138741600003942 took her fears seriously.” Exploring the experiences of family members Gupta, M., & Bowie, C. R. (2018). Family cohesion and flexibility in early of individuals at-risk of developing psychosis over 12-months. Clinical episode psychosis. Early Intervention in Psychiatry, 12(5), 886–892. Psychology & Psychotherapy, 27(1–12), 965–976. https://doi.org/10. https://doi.org/10.1111/eip.12384 1002/cpp.2483 Gurak, K., & Weisman de Mamani, A. (2016). Risk and protective factors, Jansen, J. E., Gleeson, J., & Cotton, S. (2015). Towards a better under- perceptions of family environment, ethnicity, and schizophrenia symp- standing of caregiver distress in early psychosis: A systematic review toms. Journal of Nervous and Mental Disease, 204(8), 570–577. https:// of the psychological factors involved. Clinical Psychology Review, 35, doi.org/10.1007/s11065-015-9294-9 59–66. https://doi.org/10.1016/j.cpr.2014.12.002 Haidl, T., Rosen, M., Schultze-Lutter, F., Nieman, D., Eggers, S., Jansen, J. E., Haahr, U. H., Harder, S., Trauelsen, A. M., Lyse, H. G., Heinimaa, M., Juckel, G., Heinz, A., Morrison, A., Linszen, D., Pedersen, M. B., & Simonsen, E. (2015). Caregiver distress in first- Salokangas, R., Klosterkötter, J., Birchwood, M., Patterson, P., episode psychosis: The role of subjective appraisal, over-involvement Ruhrmann, S., & European Prediction of Psychosis Study (EPOS) and symptomatology. Social Psychiatry and Psychiatric Epidemiology, 50 Group. (2018). Expressed emotion as a predictor of the first psychotic (3), 371–378. https://doi.org/10.1007/s00127-014-0935-8 episode—Results of the European prediction of psychosis study. Jungbauer, J., & Angermeyer, M. C. (2002). Living with a schizophrenic Schizophrenia Research, 199, 346–352. https://doi.org/10.1016/j. patient: A comparative study of burden as it affects parents and schres.2018.03.019 spouses. Psychiatry, 65, 110–123. https://doi.org/10.1521/psyc.65.2. Hamaie, Y., Ohmuro, N., Katsura, M., Obara, C., Kikuchi, T., Ito, F., 110.19930 Miyakoshi, T., Matsuoka, H., & Matsumoto, K. (2016). Criticism and Kavanagh, D. J. (1992). Recent developments in expressed emotion and depression among the caregivers of at-risk mental state and first- schizophrenia. British Journal of Psychiatry, 160(May), 601–620. episode psychosis patients. PLoS ONE, 11(2), 1–12. https://doi.org/10. https://doi.org/10.1192/bjp.160.5.601 1371/journal.pone.0149875 King, S. (2000). Is expressed emotion cause or effect in the mothers of Heikkilä, J., Karlsson, H., Taiminen, T., Lauerma, H., Ilonen, T., schizophrenic young adults? Schizophrenia Research, 45(1–2), 65–78. Leinonen, K. M., Wallenius, E., Virtanen, H., Heinimaa, M., https://doi.org/10.1016/S0920-9964(99)00174-7 Koponen, S., Jalo, P., Kaljonen, A., & Salakangas, R. K. R. (2002). Koutra, K., Triliva, S., Roumeliotaki, T., Basta, M., Lionis, C., & Expressed emotion is not associated with disorder severity in first- Vgontzas, A. N. (2016). Family functioning in first-episode and chronic episode mental disorder. Psychiatry Research, 111(2–3), 155–165. psychosis: The role of Patient's symptom severity and psychosocial https://doi.org/10.1016/S0165-1781(02)00134-8 functioning. Community Mental Health Journal, 52(6), 710–723. Hinojosa-Marqués, L. (2018). Family environment in early psychosis. In https://doi.org/10.1007/s10597-015-9916-y search of the psychological mechanisms underlying the manifestation Koutra, K., Triliva, S., Roumeliotaki, T., Basta, M., Simos, P., Lionis, C., & of expressed emotion in at-risk and onset stages of psychosis Vgontzas, A. N. (2015). Impaired family functioning in psychosis and (Doctoral dissertation, Universitat Autònoma de Barcelona). its relevance to relapse: A two-year follow-up study. Comprehensive Hinojosa-Marqués, L., Domínguez-Martínez, T., Cristo  bal-Narváez, P., Psychiatry, 62, 1–12. https://doi.org/10.1016/j.comppsych.2015. Kwapil, T. R., & Barrantes-Vidal, N. (2019). Ecological validity of 06.006 expressed emotion in early psychosis. Frontiers in Psychiatry, 10, 854. Koutra, K., Vgontzas, A. N., Lionis, C., & Triliva, S. (2014). Family function- https://doi.org/10.3389/fpsyt.2019.00854 ing in first-episode psychosis: A systematic review of the literature. Hinojosa-Marqués, L., Domínguez-Martínez, T., Kwapil, T. R., & Barrantes- Social Psychiatry and Psychiatric Epidemiology, 49(7), 1023–1036. Vidal, N. (2020). Predictors of criticism and emotional over- https://doi.org/10.1007/s00127-013-0816-6 involvement in relatives of early psychosis patients. PLoS ONE, 15(6), Kuipers, E., Bebbington, P., Dunn, G., Fowler, D., Freeman, D., Watson, P., e0234325. https://doi.org/10.1371/journal.pone.0234325 Hardy, A., & Garety, P. (2006). Influence of carer expressed emotion Hinojosa-Marqués, L., Domínguez-Martínez, T., Sheinbaum, S., Cristo  bal- and affect on relapse in non-affective psychosis. British Journal of Narváez, P., Kwapil, T. R., & Barrantes-Vidal, N. (2019). Relatives' Psychiatry, 188, 173–179. https://doi.org/10.1192/bjp.bp.104. attachment anxiety mediates the association between perceived loss 007294 and expressed emotion in early psychosis. PLoS ONE, 14(10), Landa, Y., Mueser, K. T., Wyka, K. E., Shreck, E., Jespersen, R., e0223425. https://doi.org/10.1371/journal.pone.0223425 Jacobs, M. A., Griffin, K. W., Van Der Gaag, M., Reyna, V. F., Hooley, J. M. (2007). Expressed emotion and relapse of psychopathology. Beck, A. T., Silbersweig, D. A., & Walkup, J. T. (2016). Development of Annual Review of Clinical Psychology, 3(1), 329–352. https://doi.org/ a group and family-based cognitive behavioral therapy program for 10.1146/annurev.clinpsy.2.022305.095236 youth at risk for psychosis. Early Intervention in Psychiatry, 10(6), Hooley, J. M., & Gotlib, I. H. (2000). A diathesis-stress conceptualization of 511–521. https://doi.org/10.1111/eip.12204 expressed emotion and clinical outcome. Applied and Preventive Psy- Law, H., Izon, E., Au-Yeung, K., Morrison, A. P., Byrne, R., Notley, C., chology, 9, 135–151. https://doi.org/10.1016/S0962-1849(05) Yung, A., Norrie, J., & French, P. (2021). Combined individual and fam- 80001-0 ily therapy in comparison to treatment as usual for people at-risk of Huang, X. Y., Sun, F. K., Yen, W. J., & Fu, C. M. (2008). The coping experi- psychosis: A feasibility study (IF CBT): Trial rationale, methodology ences of carers who live with someone who has schizophrenia. Journal and baseline characteristics. Early Intervention in Psychiatry, 15(1), of Clinical Nursing, 17(6), 817–826. https://doi.org/10.1111/j.1365- 140–148. https://doi.org/10.1111/eip.12922 2702.2007.02130.x Lee, G., Barrowclough, C., & Lobban, F. (2014). Positive affect in the family Izon, E., Berry, K., Law, H., Au-Yeung, K., & French, P. (2019). “I don't know environment protects against relapse in first-episode psychosis. Social how to fix it and sometimes it's so overwhelming” Identifying the bar- Psychiatry and Psychiatric Epidemiology, 49(3), 367–376. https://doi. riers and facilitators for family caregivers supporting someone at high- org/10.1007/s00127-013-0768-x risk of psychosis: A qualitative study. Psychosis, 12(1), 57–67. https:// Marom, S., Munitz, H., Jones, P. B., Weizman, A., & Hermesh, H. (2005). doi.org/10.1080/17522439.2019.1688858 Expressed emotion: Relevance to rehospitalization in schizophrenia Izon, E., Berry, K., Law, H., & French, P. (2018). Expressed emotion (EE) in over 7 years. Schizophrenia Bulletin, 31(3), 751–758. https://doi.org/ families of individuals at-risk of developing psychosis: A systematic 10.1093/schbul/sbi016 30  ET AL. HINOJOSA-MARQUES McFarlane, W. R., & Cook, W. L. (2007). Family expressed emotion prior to O'Brien, M. P., Miklowitz, D. J., & Cannon, T. D. (2015). Decreases in per- onset of psychosis. Family Process, 46(2), 185–197. https://doi.org/10. ceived maternal criticism predict improvement in subthreshold psy- 1111/j.1545-5300.2007.00203.x chotic symptoms in a randomized trial of family-focused therapy for McFarlane, W. R., Levin, B., Travis, L., Lucas, F. L., Lynch, S., Verdi, M., … individuals at clinical high risk for psychosis. The Journal of Family Psy- Cornblatt, B. (2015). Clinical and functional outcomes after 2 years in chology, 29(6), 945–951. https://doi.org/10.1016/j.coviro.2015.09. the early detection and intervention for the prevention of psychosis 001.Human multisite effectiveness trial. Schizophrenia Bulletin, 41(2), 532–543. O'Brien, M. P., Zinberg, J. L., Bearden, C. E., Lopez, S. R., Kopelowicz, A., https://doi.org/10.1093/schbul/sbu219 Daley, M., & Cannon, T. D. (2008). Parent attitudes and parent adoles- McGlashan, T., Miller, T., Woods, S. W., Hoffman, R., & Davidson, L. cent interaction in families of youth at risk for psychosis and with (2001). Instrument for the assessment of prodromal symptoms and recent-onset psychotic symptoms. Early Intervention in Psychiatry, 2(4), states. In T. Miller, et al. (Eds.), Early intervention in psychotic disorders 268–276. https://doi.org/10.1111/j.1751-7893.2008.00088.x (pp. 135–139). Kluwer Academic Publishers. https://doi.org/10.1007/ O'Brien, M. P., Zinberg, J. L., Ho, L., Rudd, A., Kopelowicz, A., Daley, M., 978-94-010-0892-1_7 Bearden, C. E., & Cannon, T. D. (2009). Family problem solving interac- McGorry, P. D., & Yung, A. R. (2003). Early intervention in psychosis: An tions and 6-month symptomatic and functional outcomes in youth at overdue reform. Australian and New Zealand Journal of Psychiatry, 37, ultra-high risk for psychosis and with recent onset psychotic 393–398. https://doi.org/10.1046/j.1440-1614.2003.01192.x PMID: symptoms: A longitudinal study. Schizophrenia Research, 107(2–3), 12873322 198–205. https://doi.org/10.1016/j.schres.2008.10.008 McNab, C., Haslam, N., & Burnett, P. (2007). Expressed emotion, attribu- O'Driscoll, C., Sener, S. B., Angmark, A., & Shaikh, M. (2019). Caregiving tions, utility beliefs, and distress in parents of young people with first processes and expressed emotion in psychosis, a cross-cultural, meta- episode psychosis. Psychiatry Research, 151(1–2), 97–106. https://doi. analytic review. Schizophrenia Research, 208, 8–15. https://doi.org/10. org/10.1016/j.psychres.2006.08.004 1016/j.schres.2019.03.020 Meneghelli, A., Alpi, A., Pafumi, N., Patelli, G., Preti, A., & Cocchi, A. (2011). Patterson, P. (2013). Attachment, loss and expressed emotion: Develop- Expressed emotion in first-episode schizophrenia and in ultra high-risk mental processes in psychosis. In A. Gumley, A. Gillham, K. Taylor, & patients: Results from the Programma2000 (Milan, Italy). Psychiatry M. Schwannauer (Eds.), Psychosis and emotion: The role of emotions in Research, 189(3), 331–338. https://doi.org/10.1016/j.psychres.2011. understanding psychosis, therapy and recovery (pp. 136–148). 03.021 Routledge. Miklowitz, D. J. (2004). The role of family systems in severe and recurrent Patterson, P., Birchwood, M., & Cochrane, R. (2005). Expressed emotion as psychiatric disorders: A developmental psychopathology view. Devel- an adaptation to loss: Prospective study in first-episode psychosis. opment and Psychopathology, 16(3), 667–688. https://doi.org/10. British Journal of Psychiatry, 187, 59–64. https://doi.org/10.1192/bjp. 1017/S0954579404004729 187.48.s59 Miklowitz, D. J., O'Brien, M. P., Schlosser, D. A., Addington, J., Patterson, P. J., Birchwood, M., & Cochrane, R. (2000). Preventing the Candan, K. A., Marshall, C., Domingues, I., Walsh, B. C., Zinberg, J. L., entrenchment of high Expressed Emotion in first episode psychosis: De Silva, S. D., Friedman-Yakoobian, M., & Cannon, T. D. (2014). Fam- The role of relatives grief. Australian and New Zealand Journal of Psychi- ily-focused treatment for adolescents and young adults at high risk for atry, 34 (Suppl., S191–S192), S191–S197. https://doi.org/10.1046/j. psychosis: Results of a randomized trial. Journal of the American Acad- 1440-1614.2000.00796.x emy of Child & Adolescent Psychiatry, 53(8), 848–858. https://doi.org/ Peh, O. H., Rapisarda, A., & Lee, J. (2020). Quality of parental bonding is 10.1016/j.jaac.2014.04.020 associated with symptom severity and functioning among individuals Miller, T. J., McGlashan, T. H., Rosen, J. L., Cadenhead, K., Cannon, T., at ultra-high risk for psychosis. Schizophrenia Research, 215, 204–210. Ventura, J., McFarlane, W., Perkins, D. O., Pearlson, G. D., & https://doi.org/10.1016/j.schres.2019.10.029 Woods, S. W. (2003). Prodromal assessment with the structured inter- Penn, D. L., Waldheter, E. J., Perkins, D. O., Mueser, K. T., & view for prodromal syndromes and the scale of prodromal symptoms: Lieberman, J. A. (2005). Psychosocial treatment for first-episode psy- Predictive validity, interrater reliability, and training to reliability. chosis: A research update. American Journal of Psychiatry, 162(12), Schizophrenia Bulletin, 29(4), 703–715. https://doi.org/10.1093/ 2220–2232. https://doi.org/10.1176/appi.ajp.162.12.2220 oxfordjournals.schbul.a007040 Phillips, M. R., West, C. L., Shen, Q., & Zheng, Y. (1998). Comparison of Mo, F. Y., Chung, W., Wong, S., Chun, D. Y., & Wong, K. (2007). Expressed schizophrenic patients' families and normal families in China, using emotion in relatives of Chinese patients with first-episode psychosis in Chinese versions of FACES-II and the family environment scales. Fam- Hong Kong. Hong Kong Journal of Psychiatry, 17, 38–44. ily Process, 37(1), 95–106. https://doi.org/10.1111/j.1545-5300.1998. Nordentoft, M., Thorup, A., Petersen, L., Øhlenschlæger, J., Melau, M., 00095.x Christensen, T. Ø., Krarup, G., Jørgensen, P., & Jeppesen, P. (2006). Raune, D., Kuipers, E., & Bebbington, P. E. (2004). Expressed emotion at Transition rates from schizotypal disorder to psychotic disorder for first-episode psychosis: Investigating a carer appraisal model. British first-contact patients included in the OPUS trial. A randomized clinical Journal of Psychiatry, 184, 321–326. https://doi.org/10.1192/bjp.184. trial of integrated treatment and standard treatment. Schizophrenia 4.321 Research, 83(1), 29–40. https://doi.org/10.1016/j.schres.2006.01.002 Ruhrmann, S., Schultze-Lutter, F., Salokangas, R., Heinimaa, M., O'Brien, M. P., Gordon, J. L., Bearden, C. E., Lopez, S. R., Kopelowicz, A., & Linszen, D., Dingemans, P., Birchwood, M., Patterson, P., Juckel, G., Cannon, T. D. (2006). Positive family environment predicts improve- Heinz, A., Morrison, A., Lewis, S., Graf von Reventlow, H., & ment in symptoms and social functioning among adolescents at Klosterkötter, J. (2010). Prediction of psychosis in adolescents and imminent risk for onset of psychosis. Schizophrenia Research, 81(2–3), young adults at high risk. Archives of General Psychiatry, 67, 241–251. 269–275. https://doi.org/10.1016/j.schres.2005.10.005 https://doi.org/10.1001/archgenpsychiatry.2009.206 O'Brien, M. P., Miklowitz, D. J., Candan, K. A., Marshall, C., Domingues, I., Salinger, J. M., O'Brien, M. P., Miklowitz, D. J., Marvin, S. E., & Walsh, B. C., Zinberg, J. L., de Silva, S. D., Woodberry, K. A., & Cannon, T. D. (2018). Family communication with teens at clinical Cannon, T. D. (2014). A randomized trial of family focused therapy high-risk for psychosis or bipolar disorder. Journal of Family Psychology, with populations at clinical high risk for psychosis: Effects on interac- 32, 507–516. https://doi.org/10.1037/fam0000393 tional behavior. Journal of Consulting and Clinical Psychology, 82(1), Santesteban-Echarri, O., MacQueen, G., Goldstein, B. I., Wang, J., 90–101. https://doi.org/10.1037/a0034667 Kennedy, S. H., Bray, S., Lebel, C., & Addington, J. (2018). Family  ET AL. HINOJOSA-MARQUES 31 functioning in youth at-risk for serious mental illness. Comprehensive Psychology, 15, 157–165. https://doi.org/10.1111/j.2044-8260.1976. Psychiatry, 87, 17–24. https://doi.org/10.1016/j.comppsych.2018. tb00021.x 08.010 Wang, L., Shi, J., Chen, F., Yao, Y., Zhan, C., Yin, X., Fang, X. Y., Wang, H. J., Schlosser, D. A., Pearson, R., Perez, V. B., & Loewy, R. L. (2012). Environ- Yuan, J. B., & Zhao, X. (2015). Family perception and 6-month mental risk and protective factors and their influence on the emer- symptomatic and functioning outcomes in young adolescents at clini- gence of psychosis. Adolescent Psychiatry, 2(2), 163–171. https://doi. cal high risk for psychosis in a general population in China. PLoS ONE, org/10.2174/2210676611202020163 10(9), 1–13. https://doi.org/10.1371/journal.pone.0138361 Schlosser, D. A., Zinberg, J. L., Loewy, R. L., Casey-Cannon, S., Wearden, A. J., Tarrier, N., Barrowclough, C., Zastowny, T. R., & O'Brien, M. P., Bearden, C. E., Vinogradov, S., & Cannon, T. D. (2010). Rahill, A. A. (2000). A review of expressed emotion research in health Predicting the longitudinal effects of the family environment on pro- care. Clinical Psychology Review, 20(5), 633–666. https://doi.org/10. dromal symptoms and functioning in patients at-risk for psychosis. 1016/S0272-7358(99)00008-2 Schizophrenia Research, 118(1–3), 69–75. https://doi.org/10.1016/j. Weisman, A. (2005). Integrating culturally based approaches with existing schres.2010.01.017 interventions for Hispanic/Latino families coping with schizophrenia. Smith, L. M., Onwumere, J., Craig, T., & Kuipers, E. (2018). Caregiver corre- Psychotherapy, 42(2), 178–197. https://doi.org/10.1037/0033-3204. lates of patient-initiated violence in early psychosis. Psychiatry 42.2.178 Research, 270, 412–417. https://doi.org/10.1016/j.psychres.2018. Welsh, P., & Tiffin, P. A. (2015). Adolescent family perceptions in the At- 09.011 Risk Mental State for psychosis. Early Intervention in Psychiatry, 9(4), Strachan, A. M., Feingold, D., Goldstein, M. J., Miklowitz, D. J., & 316–323. https://doi.org/10.1111/eip.12115 Nuechterlein, K. H. (1989). Is expressed emotion an index of a transac- Wiedemann, G., Rayki, O., Feinstein, E., & Hahlweg, K. (2002). The Family tional process? II. Patient's coping style. Family Process, 28, 169–181. Questionnaire: Development and validation of a new self-report scale https://doi.org/10.1111/j.1545-5300.1989.00169.x for assessing expressed emotion. Psychiatry research, 109(3), 265–279. Thompson, E., Rakhshan, P., Pitts, S. C., Demro, C., Millman, Z. B., https://doi.org/10.1016/S0165-1781(02)00023-9 Bussell, K., DeVylder, J., Kline, E., Reeves, G. M., & Schiffman, J. Wong, C., Davidson, L., McGlashan, T., Gerson, R., Malaspina, D., & (2019). Family functioning moderates the impact of psychosis-risk Corcoran, C. (2008). Comparable family burden in families of clinical symptoms on social and role functioning. Schizophrenia Research, 204 high-risk and recent-onset psychosis patients. Early Intervention in Psy- (337–342), 6–11. https://doi.org/10.1016/j.schres.2018.08.035 chiatry, 2(4), 256–261. https://doi.org/10.1111/j.1751-7893.2008. Tomlinson, E., Onwumere, J., & Kuipers, E. (2014). Distress and negative 00086.x experiences of the caregiving relationship in early psychosis: Does Yee, C. I., Gupta, T., Mittal, V. A., & Haase, C. M. (2020). Coping with family social cognition play a role? Early Intervention in Psychiatry, 8(3), 253– stress in individuals at clinical high-risk for psychosis. Schizophrenia 260. https://doi.org/10.1111/eip.12040 Research, 216, 222–228. https://doi.org/10.1016/j.schres.2019.  pez, S., Marvin, S., Zinberg, J., Cannon, T. D., O'Brien, M., & Tsai, K. H., Lo 11.057 Bearden, C. E. (2015). Perceptions of family criticism and warmth and Yung, A. R., Phillips, L. J., Yuen, H. P., & McGorry, P. D. (2004). Risk factors their link to symptom expression in racially/ethnically diverse adoles- for psychosis in an ultra high-risk group: Psychopathology and clinical cents and young adults at clinical high risk for psychosis. Early Interven- features. Schizophrenia Research, 67(2–3), 131–142. https://doi.org/ tion in Psychiatry, 9(6), 476–486. https://doi.org/10.1111/eip.12131 10.1016/S0920-9964(03)00192-0 Vasconcelos e Sa, D., Wearden, A., & Barrowclough, C. (2013). Expressed emotion, types of behavioral control and controllability attributions in relatives of people with recent-onset psychosis. Social Psychiatry and Psychiatric Epidemiology, 48(9), 1377–1388. https://doi.org/10.1007/ How to cite this article: Hinojosa-Marqués, L., Domínguez- s00127-013-0659-1 Martínez, T., & Barrantes-Vidal, N. (2021). Family Vasconcelos e Sa, D., Wearden, A., Hartley, S., Emsley, R., & Barrowclough, C. (2016). Expressed Emotion and behaviorally control- environmental factors in at-risk mental states for psychosis. ling interactions in the daily life of dyads experiencing psychosis. Psy- Clinical Psychology & Psychotherapy, 1–31. https://doi.org/10. chiatry Research, 245, 406–413. https://doi.org/10.1016/j.psychres. 1002/cpp.2651 2016.08.060 Vaughn, C., & Leff, J. (1976). The measurement of expressed emotion in the families of psychiatric patients. British Journal of Social & Clinical