Internal structure of the Community Assessment of Psychic Experiences—Positive (CAPE-P15) scale: Evidence for a general factor
Introduction
Psychotic-like experiences (PLEs) are common in the general population (Scott et al., 2008), particularly during childhood and adolescence (Laurens et al., 2012), with a rapid decline in prevalence with age (Rössler et al., 2007). The median prevalence of PLEs is 17% in children aged 9 to 12 and 7.5% in adolescents aged 13 to 18 years (Kelleher et al., 2012a, Kelleher et al., 2012b), but marked variations across countries have been found (Nuevo et al., 2010). PLEs are regarded as risk factors for future psychotic disorders (Hanssen et al., 2005, Kelleher et al., 2012a) and have been proposed as important markers to identify at-risk subjects (Kelleher and Cannon, 2011). However, not all individuals experiencing PLEs will develop clinical psychosis (Kline et al., 2014), and these experiences may represent normal aspects of the growth process (Wigman et al., 2011a, Wigman et al., 2011b). The risk of later psychotic outcomes may increase when PLEs are persistent over time (Cougnard et al., 2007, Domínguez et al., 2011), when they interact with additional risk factors, and/or when protective factors are absent (Mackie et al., 2011, Wigman et al., 2011a). However, it has been posited that PLEs may not be pathognomonic for psychotic disorders (Fisher et al., 2013), although they are associated with poor mental health (Nuevo et al., 2010), higher sensitivity to daily life stress (Collip et al., 2013), cannabis use (Stefanis et al., 2004), depressive symptoms (Barragan et al., 2011), distress, depression and poor functioning (Armando et al., 2010), and risk of suicide attempts in both general adolescent samples (Kelleher et al., 2013, Saha et al., 2011) and help-seeking adolescent subjects (Kelleher et al., 2014, Nishida et al., 2014).
Based on previous findings demonstrating strong associations between PLEs and psychiatric disorders, and given the evidence showing the potential role of PLEs as predictors of psychosis onset in adolescents and young adults, there is increasing interest in elucidating their underlying structure, which remains debated (Therman et al., 2014). According to the multidimensional perspective, PLEs are heterogeneous and differentially associated with mental health problems (Armando et al., 2010, Barragan et al., 2011). However, it has also been posited that these experiences could be represented by a unidimensional construct, as Laurens et al. (2012) observed in a large sample of children aged 9–11 years, which parallels recent findings showing the existence of a general psychosis dimension in both early psychosis and enduring psychosis (Reininghaus et al., 2013).
Various self-report screening questionnaires have been developed to measure PLEs in the general population (Kline et al., 2012). The use of these questionnaires, combined with additional instruments that address general psychopathology, may facilitate the identification of adolescents who need clinical evaluation and/or intervention (Kelleher et al., 2011). A widely used self-report instrument to address PLEs is the Community Assessment of Psychic Experiences (CAPE). Its original version (42 items, three subscales: positive, depressive and negative symptoms) (Stefanis et al., 2002) has high internal reliability and construct and discriminant validity (Konings et al., 2006, Brenner et al., 2007). The CAPE is regarded as an appropriate tool to detect individuals at an increased risk of developing psychosis in clinical and help-seeking populations (Mossaheb et al., 2012). Prior studies using the 20-item positive subscale (CAPE-P) have found a similar and stable structure with 3–5 factors, revealing that specific subtypes of PPEs are differentially related to distress, depression and poor functioning (Armando et al., 2010, Barragan et al., 2011, Wigman et al., 2012). A recent study investigated the internal structure of the CAPE-P and found that a three-factor model yielded the best fit when items addressing magical thinking, grandiosity, and paranormal beliefs were omitted (Capra et al., 2013). The proposed factors were bizarre experiences (BEs), persecutory ideation (PI) and perceptual abnormalities (PAs), which resulted in a new 15-item version of the questionnaire (CAPE-P15). This factorial solution solves problems detected by prior research, mainly those related to the excluded items (misinterpretation, cross-loading and unclear associability with psychopathology) and fits well with previous findings showing strong associations between the proposed factors and mental health problems (Yung et al., 2009, Wigman et al., 2011a, Wigman et al., 2011b).
The purpose of the present study is two-fold. First, we seek to provide further evidence on the internal structure of the CAPE-P15 by conducting exploratory and confirmatory analyses to contrast two possible factor solutions. Second, we attempt to expand the scope of the research on and use of the CAPE-P15 from the population studied by Capra et al. (2013) (English-speaking university students, mostly females) to a broader population of Spanish-speaking adolescents aged 13–18.
We believe that the results of this study yield new insights into the underlying positive features of PLEs and enhance the range of application of the instrument.
Section snippets
Participants
A cross-sectional study was conducted with 727 high school students (adolescents) (women = 50.7%) aged 13 to 18 years (mean = 15.4, ±1.33), recruited between April and July 2014 in six secondary schools of the city of Talca, Chile. Only 2 subjects rejected to participate in the study, and then, nearly 100% of the subjects attending to each school were recruited. We observed that 3.9% of subjects have one or more missing values, with a distribution completely at random (MCAR, Little's test sig. =
ESEM analysis to test a set of interpretable models
The goodness-of-fit level was unacceptable for the unidimensional model (A), moderate for the two-factor model (B) and good for the three-factor model (C) (Supplementary Table 2). The moderate to high correlations observed among the factors of model C suggest a possible superior order structure. Consequently, we fitted a hierarchical ESEM model (D) with three specific factors and a general factor, which yielded an excellent fit (CFI = 0.99, TLI = 0.99, RMSEA = 0.025) (range of G factor loadings:
Discussion
We examined the internal structure of positive psychotic experiences addressed by the CAPE-P15 (Capra et al., 2013) in a sample of non-help-seeking adolescents aged 13–18 years. The comparison of different competing models yielded two latent structures with acceptable fit levels: a three-correlated-factor model (persecutory ideation, bizarre experiences and perceptual abnormalities) and a hierarchical model (one general PLE factor and three specific factors). Our initial analyses replicated the
Role of funding source
This work was supported by the Dirección de Investigación, Universidad de Talca (I002981).
Contributors
Daniel Nunez designed the study and directed its implementation, did the literature searches, and wrote the manuscript.
Victor Arias performed the statistical analyses, and wrote the manuscript.
Edgar Vogel and Gómez reviewed the manuscript and revised critically for intellectual content.
Conflict of interest
None of the authors have any conflicts of interest.
Acknowledgments
The authors would like to thank all the students for their participation.
References (66)
- et al.
Psychotic-like experiences and correlation with distress and depressive symptoms in a community sample of adolescents and young adults
Schizophr. Res.
(2010) - et al.
Do different psychotic experiences differentially predict need for care in the general population?
Compr. Psychiatry
(2005) - et al.
Psychotic-like experiences and depressive symptoms in a community sample of adolescents
Eur. Psychiatry
(2011) - et al.
Validation of the English and French versions of the Community Assessment of Psychic Experiences (CAPE) with a Montreal community sample
Schizophr. Res.
(2007) - et al.
Brief screening for psychosis-like experiences
Schizophr. Res.
(2013) - et al.
Convergent and discriminant validity of attenuated psychosis screening tools
Schizophr. Res.
(2012) - et al.
Psychosis-like experiences and distress among adolescents using mental health services
Schizophr. Res.
(2014) - et al.
Development of a brief self-report questionnaire for screening putative pre-psychotic states
Schizophr. Res.
(2013) - et al.
The structure of negative emotional states: comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories
Behav. Res. Ther.
(1995) - et al.
Short scales for measuring schizotypy
Schizophr. Res.
(2005)