Elsevier

Schizophrenia Research

Volume 150, Issue 1, October 2013, Pages 197-204
Schizophrenia Research

Deconstructing sub-clinical psychosis into latent-state and trait variables over a 30-year time span

https://doi.org/10.1016/j.schres.2013.07.042Get rights and content

Abstract

Background

Our aim was to deconstruct the variance underlying the expression of sub-clinical psychosis symptoms into portions associated with latent time-dependent states and time-invariant traits.

Methods

We analyzed data of 335 subjects from the general population of Zurich, Switzerland, who had been repeatedly measured between 1979 (age 20/21) and 2008 (age 49/50). We applied two measures of sub-clinical psychosis derived from the SCL-90-R, namely schizotypal signs (STS) and schizophrenia nuclear symptoms (SNS). Variance was decomposed with latent state–trait analysis and associations with covariates were examined with generalized linear models.

Results

At ages 19/20 and 49/50, the latent states underlying STS accounted for 48% and 51% of variance, whereas for SNS those estimates were 62% and 50%. Between those age classes, however, expression of sub-clinical psychosis was strongly associated with stable traits (75% and 89% of total variance in STS and SNS, respectively, at age 27/28). Latent states underlying variance in STS and SNS were particularly related to partnership problems over almost the entire observation period. STS was additionally related to employment problems, whereas drug-use was a strong predictor of states underlying both syndromes at age 19/20. The latent trait underlying expression of STS and SNS was particularly related to low sense of mastery and self-esteem and to high depressiveness.

Conclusions

Although most psychosis symptoms are transient and episodic in nature, the variability in their expression is predominantly caused by stable traits. Those time-invariant and rather consistent effects are particularly influential around age 30, whereas the occasion-specific states appear to be particularly influential at ages 20 and 50.

Introduction

In the past two decades, several studies have demonstrated that the expression of a psychosis phenotype can be observed below the threshold of its clinical detection (van Os et al., 2000, Wiles et al., 2006, Rössler et al., 2007). The occurrence of psychotic symptoms in the general population can be characterized as a continuum with differing levels of severity and persistence (Rössler et al., 2013). Van Os et al. (2009) have found in their systematic review that the median prevalence is approximately 5% for sub-clinical psychosis, which is at least five-fold higher than the prevalence for diagnosed schizophrenia (Rössler et al., 2005) or three to four times higher for non-affective psychosis in the general population (Kendler et al., 1996, Perala et al., 2007).

Van Os' research group has estimated that 75–90% of those sub-clinical psychosis symptoms are transitory and disappear over time. Otherwise we could demonstrate, that sub-clinical psychosis symptoms are quite persistent over time in some individuals (Rössler et al., 2007). Thus, subclinical psychosis may as well indicate a more stable underlying psychopathology. This latter assumption is in agreement with the concept of schizotypal personality disorder, which has been defined as a stable maladaptive personality trait (American Psychiatric Association, 2000).

There are several theoretical models that describe how sub-clinical psychosis symptoms might arise and persist. Psychosis symptoms might express an underlying liability. Such an underlying liability is not restricted to psychosis symptoms and can provoke all kind of transient psychopathological symptoms (Rössler et al., 2011b). And psychosis symptoms can also be triggered by environmental influences (for instance by acute stress). Finally, those (psychosis) symptoms provoked by the social environment can interact with various other individual personality dimensions, which can alternatively ameliorate or deteriorate the clinical picture. As a result the affected subjects then either recover more quickly or develop more enduring psychopathological manifestations.

We are not aware of any study that has attempted to determine the longitudinal latent state–trait structure of sub-clinical psychosis. To date, we are still uncertain whether liability to sub-clinical psychosis represents either transient and occasion-specific states or a stable dispositional trait. Furthermore, the particular proportions of states and trait might vary over time. Thus, this study is the first to specifically analyze the latent state–trait structure of sub-clinical psychosis within a community sample that entails a cohort of adults evaluated seven times between the ages of 20 and 50.

Our study objectives were to: i) determine the proportion of variance explained in subclinical psychosis related to latent states and trait over a 30-year time span within a community sample, and ii) identify coping strategies, personality dimensions or environmental factors, which might relate to these latent states and trait.

Section snippets

Sampling procedure

The Zurich Study comprised a cohort of 4547 subjects (m = 2201; f = 2346) representative of the canton of Zurich in Switzerland, who were screened in 1978 with the Symptom Checklist 90 Revised (SCL-90-R) (Derogatis, 1977) when males were 19 and females were 20 years old. A stratified subsample of those participants was selected for comprehensive face-to-face interviews and subsequent follow-ups. Such a two-phase procedure is fairly common in epidemiological research (Dunn et al., 1999) and is

Results

We first inspected the fit of our LST models as indicated in Fig. 1, Fig. 2. For the STS model the χ2 was 114.464 (df = 63), p < 0.001, the CFI was 0.977, the TLI was 0.967, the RMSEA was 0.049, and the SRMR was 0.041. Except for the χ2-test these indices demonstrated a good model fit for STS. With respect to the SNS model the χ2 was 70.840 (df = 63), p = 0.233, the CFI was 0.989, the TLI was 0.984, the RMSEA was 0.019, and the SRMR was 0.041, which indicates an excellent model fit. The reliability

Discussion

To the best of our knowledge, this is the first study to determine the proportion of variance explained in sub-clinical psychosis related to latent states and trait characteristics over a 30-year span. We analyzed data from the prospective Zurich cohort study and fitted a latent state–trait (LST) model for two subscales derived from the well-established SCL-90-R. A latent state typically comprises transient effects specifically related to a particular measurement occasion, such as age-related

Role of funding source

This work was supported by the Swiss National Science Foundation (Grant Number 32-50881.97). The sponsor played no role in the design or conduct of the study; the collection, management, analysis, or interpretation of data; the preparation, review, and approval of the manuscript; or the decision to submit the manuscript for publication.

Contributors

Wulf Rössler took responsibility for all assessment waves since 1999, and also drafted and revised the manuscript. Michael Hengartner conducted all statistical analyses and substantially contributed to drafting and critical revision of the manuscript. Vladeta Ajdacic-Gross and Helene Haker substantially contributed to drafting and critical revision of the manuscript. Jules Angst designed the study, was responsible for assessment waves until 1993, and substantially contributed to drafting and

Conflict of interest

None.

References (46)

  • J.H. Steiger

    Understanding the limitations of global fit assessment in structural equation modeling

    Pers. Individ. Dif.

    (2007)
  • J. van Os et al.

    Strauss (1969) revisited: a psychosis continuum in the general population?

    Schizophr. Res.

    (2000)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorders DSM-IV-TR

    (2000)
  • J. Angst et al.

    The Zurich Study—a prospective epidemiological study of depressive, neurotic and psychosomatic syndromes. I. Problem, methodology

    Eur. Arch. Psychiatry Neurol. Sci.

    (1984)
  • J. Angst et al.

    Prevalence of mental disorders in the Zurich Cohort Study: a twenty year prospective study

    Epidemiol. Psichiatr. Soc.

    (2005)
  • E. Barkus et al.

    Substance use in adolescence and psychosis: clarifying the relationship

    Annu. Rev. Clin. Psychol.

    (2010)
  • L.R. Derogatis

    Symptom Checklist 90, R-Version Manual I: Scoring, Administration, and Procedures for the SCL-90

    (1977)
  • G. Dunn et al.

    Two-phase epidemiological surveys in psychiatric research

    Br. J. Psychiatry

    (1999)
  • D. Eich et al.

    The Zurich Study: participation patterns and Symptom Checklist 90-R scores in six interviews, 1979–99

    Acta Psychiatr. Scand. Suppl.

    (2003)
  • M. Eid et al.

    Do you feel better or worse? The validity of perceived deviations of mood states from mood traits

    Eur. J. Pers.

    (1999)
  • J. Fahrenberg et al.

    Das Freiburger Persönlichkeitsinventar FPI

    Revidierte Fassung FPI-R und teilweise geänderte Fassung FPI-A1, 4. revidierte Auflage

    (1984)
  • C. Geiser

    Datenanalyse mit Mplus: Eine anwendungsorientierte Einführung

    (2011)
  • M.P. Hengartner et al.

    Interpersonal functioning deficits in association with DSM-IV personality disorder dimensions

    Soc. Psychiatry Psychiatr. Epidemiol

    (2013)
  • Cited by (0)

    View full text