In search of neurobiological intermediate phenotypes of
aberrant information processing in psychosis
ISBN | 978-90-8891-169-9
Cover image | brain coral, Red Sea | Victor Pardede
Cover design | Claudia Simons | Box Press B.V.
Print | Box Press B.V., Oisterwijk
© 2010 C.J.P. Simons, Maastricht. All rights reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise, without the prior permission in
writing from the proprietor.
In search of neurobiological intermediate phenotypes of
aberrant information processing in psychosis
ter verkrijging van de graad van doctor aan de Universiteit Maastricht
op gezag van de Rector Magnificus, Prof. mr. G.P.M.F. Mols,
volgens het besluit van het College van Decanen
in het openbaar te verdediging op vrijdag 18 juni om 10.00 uur
Claudia Johanna Petronella Simons
geboren op 7 september 1980 te Tegelen
Prof. dr. J. van Os
Prof. dr. W. Riedel
Dr. L. Krabbendam (Vrije Universiteit Amsterdam)
Prof. dr. M. de Vries (voorzitter)
Dr. A.L. van Bemmel (Geestelijke Gezondheidszorg Eindhoven en de Kempen)
Prof. dr. P. van Harten
Dr. I. Myin-Germeys
Prof. dr. W.M.A. Verhoeven (GGz Noord- en Midden-Limburg, Erasmus MC)
South Limburg Mental Health Research and Teaching Network, PhD Series
The research presented in this thesis was conducted at the School for Mental Health and
Neuroscience, Department of Psychiatry and Neuropsychology, Maastricht University.
The publication of this thesis was financially supported by: Eli Lilly Nederland B.V.,
Lundbeck B.V., Schering-Plough Nederland B.V., Servier Nederland Farma B.V.
Mari de Gracia Dominguez
Subclinical psychotic experiences and cognitive functioning as a bivariate
phenotype for genetic studies in the general population
Cognition as predictor of current and follow-up depressive symptoms in the
Auditory P300 and N100 components as intermediate phenotypes for
psychotic disorder: familial liability and reliability
Subtle gene-environment interactions driving paranoia in daily life
Functional magnetic resonance imaging of inner speech in schizophrenia
Curriculum vitae 137
List of publications 139
Phenomenology of psychosis
Psychotic disorders, of which schizophrenia is the most prevalent and severe, are
characterised by a marked distortion in the perception of reality. Traditionally, the main
symptoms of psychotic disorder have been subdivided into positive and negative symptoms.
Positive symptoms (e.g. hallucinations, delusions) are experiences that are not present in
healthy individuals and tend to be of a fluctuating nature. In contrast, negative symptoms (e.g.
flattening of affect, apathy, poverty of thoughts or speech) reflect an absence of certain
abilities and impulses present in healthy individuals and tend to be of a more chronic nature
(Green, 2001). Affective symptoms also form an integral part of psychotic disorders. In
addition, it has been well established that psychotic disorders are characterised by robust
neurocognitive deficits, although they are not part of the diagnostic criteria for psychosis.
Deficits can be found across a broad array of cognitive domains, such as auditory attention,
episodic memory, and working memory (Faraone, et al., 1999; Krabbendam, et al., 2001) and
have been regarded as predictors of functional outcome (Green, et al., 2004).
Psychotic symptoms have been traditionally regarded as clinical manifestations that can be
either present or absent, a dichotomy still employed in clinical practice up until this day.
However, there is accumulating evidence that the psychosis phenotype is better perceived as a
continuous rather than categorical phenotype, extending from subclinical experiences to
clinical symptoms (Johns and van Os, 2001; Stefanis, et al., 2002; van Os, et al., 2009;
Verdoux and van Os, 2002). These subclinical psychotic experiences, alternatively labeled as
schizotypy, psychosis-proneness or at-risk mental states, are prevalent in the general
population (Eaton, et al., 1991; Peters, et al., 1999; Tien, 1991; Van Os, et al., 1999).
Evidence that subclinical psychotic experiences are related to underlying aetiological
influences similar to those underlying psychotic disorder further supports a continuity model
of psychosis. Investigating psychotic experiences in individuals with non-clinical expression
of the psychosis phenotype may therefore be a fruitful approach to studying the underlying
aetiology of psychosis and has the advantage that it avoids potential confounds of illness-
related factors that do not reflect psychosis vulnerability, such as institutionalisation,
stigmatisation and antipsychotic drug treatment.
Aetiology of psychosis
Schizophrenia is a particularly heterogeneous disorder with multiple causes. It has been well
established that schizophrenia is a highly heritable disorder, with an estimated contribution of
genetic factors and gene–environment interactions to schizophrenia vulnerability of
approximately 80% (Cardno and Gottesman, 2000; Gottesman, 1991; Sullivan, et al., 2003).
However, genetic association studies have, as yet, failed to provide consistent results
regarding the precise mode of transmission of the genetic vulnerability. The complex genetic
architecture of the disorder has led to the search for intermediate phenotypes with a simpler
genetic basis than the dichotomous schizophrenia phenotype. Intermediate phenotypes,
alternatively called endophenotypes, are biological markers that: 1. are associated with the
illness in the population, 2. are heritable, 3. are state-independent, 4. cosegregate with the
illness within families, and 5. are present in unaffected relatives of patients to a higher degree
than in the general population (Gottesman and Gould, 2003). Studying such quantifiable,
relatively simpler and biologically-based intermediate phenotypes may guide genetic research
and assist in neurobiological validation of the overarching psychosis phenotype.
Neurocognitive intermediate phenotypes
Genetic mechanisms are likely to impact on developing brain systems, resulting in a
predisposition to schizophrenia and manifest across a wide range of neurocognitive faculties.
Consequently, behavioural measures of neurocognitive functioning are promising potential
intermediate phenotypes as are neurophysiological assessments of brain systems that bring
about specific neurocognitive processes. Neurocognitive deficits have been widely recognised
as core features of psychosis and have been shown i) to be heritable, ii) to cluster in families
of patients with a psychotic disorder and, iii) to be present in first-degree relatives of patients
at levels that are intermediate between healthy controls and patients (Krabbendam, et al.,
2001; Kremen, et al., 1994). These findings underline the well-recognized notion that deficits
in cognitive functioning can be useful intermediate phenotypes in studying psychosis
(Gottesman and Gould, 2003). At a behavioural level, largest effect sizes of neurocognitive
deficits have been reported for attention/information processing speed, memory, and
executive functioning (Braff, 1993; Dickinson, et al., 2007; Heinrichs and Zakzanis, 1998;
Morice and Delahunty, 1996; Nuechterlein and Dawson, 1984). At the neurophysiological
level, one of the most promising intermediate phenotypes is the P300 waveform. Amplitude
reduction of the auditory P300 wave is a robust finding in patients with a psychotic disorder
and P300 latency has been shown to be delayed in patients compared with healthy controls
(Bramon, et al., 2004; Jeon and Polich, 2001), reflecting abnormalities in attention to
incoming stimulus information when representations are updated (P300 amplitude) (Polich,
2007; Turetsky, et al., 2007) and processing speed (P300 latency) (Polich, 2007).
Gen–environment interactions: stress sensitivity, COMT and BDNF
Genetic mechanisms are likely to exert their influence on developing brain systems not in
isolation, but in symphony with environmental influences (van Os and Kapur, 2009). The
impact of environmental stressors depends on the genetic makeup of the individual and any
impact of genetic variation on the phenotype is contextualized by the environment (van Os, et
al., 2008). This is essentially what is described by the widely accepted vulnerability–stress
model (Zubin and Spring, 1977).
Aversive life events (Bebbington, et al., 1993; Bebbington, et al., 1996), victimisation and
childhood trauma (Bebbington, et al., 2004; Janssen, et al., 2004; Read, 1997), and aversive
family environments (Bebbington and Kuipers, 1994; Butzlaff and Hooley, 1998) have been
consistently linked to vulnerability for psychotic disorder. These studies show that
environmental stressors can play a significant role in the emergence of psychotic symptoms,
but these studies can not establish whether sensitivity to stress may be an intermediate
phenotype because they did not asses the individual’s way of coping with the stressors. The
Experience Sampling Method (ESM) does enable the study of moment-to-moment
interactions between daily life stressors and the individual. The ESM is a momentary
assessment technique to assess subjects in their daily living environment and has been
extensively validated for the use of immediate effects of stressors on mood (Csikszentmihalyi
and Larson, 1987; Delespaul, 1995; DeVries, 1992; Myin-Germeys, et al., 2001; Wichers, et
al., 2007). These studies have shown that stress sensitivity (defined in terms of affective or
psychotic responses to self-reported daily life stressors) is higher in patients with a psychotic
disorder than in healthy controls, with first-degree relatives showing intermediate scores for
stress sensitivity (Myin-Germeys, et al., 2005; Myin-Germeys, et al., 2001). Evidence has
also been found that this heightened stress sensitivity i) is state-independent as stress
sensitivity was investigated in patients in remission (Myin-Germeys, et al., 2005; Myin-
Germeys, et al., 2001), and ii) cosegregates within families (Lataster, et al., 2009), suggesting
stress sensitivity may be a potential intermediate phenotype.
Two promising genes that may moderate psychotic reactivity to environmental stressors are
the catechol-O-methyltransferase (COMT) and brain-derived neurotrophic factor (BDNF)
genes. The COMT gene is a biologically plausible candidate given that it codes for catechol-
O-methyltransferase, an enzyme that plays a critical role in the degradation of dopamine. The
COMT gene contains a functional polymorphism (Val158Met) with two common variants
(valine and methionine), corresponding to high and low activity, respectively. Increased
COMT activity may result in i) reduced dopamine neurotransmission in the prefrontal cortex,
and, subsequently, ii) increased levels of mesolimbic dopamine signalling (Bilder, et al.,
2004). The first has been associated with neurocognitive impairments, including impairments
in working memory, attention, and executive functioning (Bilder, et al., 2002; Egan, et al.,
2001; Nolan, et al., 2004), whereas the latter has been associated with increased risk for
delusions and hallucinations (Akil, et al., 2003; Bilder, et al., 2004).
Based on the neurodevelopmental hypothesis, BDNF is also a conceivable candidate gene
because of the acknowledged effects of BDNF on neurodevelopment and neuronal activity
(Hua and Smith, 2004; Huang, et al., 1999). A Val66Met functional polymorphism in the gene
encoding BDNF has been the focus of much recent investigation. The valine (Val) variant is
associated with higher neuronal BDNF secretory activity than the methionine (Met) variant
(Chen, et al., 2004). The polymorphism has been linked to hippocampal functioning in
humans and has correspondingly been thought to affect episodic memory functioning (Egan,
et al., 2003) and sensitivity to stress. Met carriers may be more sensitive to stress-induced
BDNF depletion and may thus experience more extreme behavioural responses to stress
compared with Val/Val subjects.
Monitoring of inner speech
Although neurocognitive intermediate phenotypes such as information processing speed and
memory have been linked to the negative symptom dimension rather than other symptom
dimensions (de Gracia Dominguez, et al., 2009), they are not linked to specific psychotic
symptoms, but rather seem to indicate a general vulnerability. Identifying cognitive
abnormalities specific to psychosis is the aim of the cognitive neuropsychiatric approach. This
approach attempts to explain psychiatric phenomena such as delusions and hallucinations in
terms of aberrations in normal cognitive mechanisms. Using a symptom-oriented approach
can greatly facilitate our understanding of underlying cognitive mechanisms and tends to
demonstrate greater reliability and biological validity than a syndrome-oriented approach.
Auditory verbal hallucinations (AVHs) are one of the most common symptoms in
schizophrenia, often described as ‘hearing voices’. One of the most influential current
cognitive models suggests that auditory hallucinations are the result of defective self-
monitoring. Self-monitoring refers to the cognitive capacity to correctly differentiate
information that is internally generated from information that is generated by an external
source (Johnson, et al., 1993). If this monitoring system is defective, verbal thoughts will not
be recognized as being self-generated, leading to the experience of AVHs (Feinberg, 1978).
The development of neuroimaging techniques has given investigators tools to study the
biological basis of cognitive processes implicated in AVHs. Neuroimaging studies have
revealed that imagining another person’s speech yields normal activation of the left inferior
frontal gyrus but abnormal activation of the temporal cortex in schizophrenia patients prone to
auditory hallucinations compared with schizophrenia patients without a history of AVHs and
healthy controls (McGuire, et al., 1996; Shergill, et al., 2000), thus adding biological
plausibility to the cognitive model.
Aims and outline of this thesis
The overall aim of this thesis was to examine dysfunctions in information processing in
relation to various symptoms of the psychosis phenotype and in relation to neurobiological
Our first aim was to investigate neurocognitive functions as potential intermediate phenotypes
for psychosis. In chapters 2 and 3, we looked at the relationship between behavioural
measures of neurocognitive functioning and symptom dimensions in a general population
twin sample. In chapter 2 we examined whether cognitive deficits are related to specific
dimensions of subclinical psychotic experiences and whether associations between these
variables are caused by additive genetic, common environmental, and/or individual-specific
environmental factors. In chapter 3 we looked at depressive symptoms, which form an
integral part of the psychosis phenotype. We examined whether poor cognitive functioning
can be a risk factor in the development of depressive symptoms. In chapter 4 we investigated
the physiological correlates of information processing mechanisms in patients with a
psychotic disorder, their first-degree relatives, and healthy controls, focussing on two
important features potential intermediate phenotype should demonstrate: familiality and
Our second aim was to investigate gene–environment interactions. In chapter 5 we examined
whether the COMT Val158Met and BDNF Val66Met polymorphisms in part mediate genetic
effects on paranoid reactivity to minor stressors, using the same general population twin
sample as in chapters 2 and 3.
Our third aim was to investigate self-monitoring of inner speech. In chapter 6 we investigated
the functional neuroanatomy of inner and external speech in both schizophrenia patients with
a history of auditory verbal hallucinations and healthy controls.
Finally, chapter 7 closes with a general discussion.