Schizophrenia is a chronic and severe mental disorder with a typical onset in adolescence and early adulthood and a lifetime prevalence of about 1%. On average, males have their illness onset 3 to 4 years earlier than females. Onset of schizophrenia is very rare before age 11, and prior to age 18 the illness has been called “early-onset schizophrenia” (EOS), while onset before age 13 has been termed “very early-onset schizophrenia” (VEOS; Werry, 1981).
Prior to examining topics in schizophrenia, we must address a basic question as to the definition of adolescents and adults. The way these groups will be defined is partly related to the question being asked. That is, research studies that emphasize the study of neural development or finding links between endocrine changes and onset of schizophrenia are likely to place more emphasis on defining adolescence in terms of body or brain maturation. For example, adolescence could be defined as the period between the onset and offset of puberty. Alternatively, it could be defined on the basis of our current knowledge of brain development, which suggests that maturational processes accelerate around the time of puberty but that they continue on into what is often considered young adulthood. Most recent studies of normal brain development suggest that brain maturation continues to the early 20s. If this rather extended definition of adolescence is used, then the appropriate adult contrast groups are likely to be somewhat older—people in their late 20s, 30s, or even 40s.
Under the general rubric of phenomenology, four major topics need to be considered as we explore the relevance of research on adults to the understanding of adolescents. These four topics are diagnostic criteria, phenomenology, the relationship of phenomenology to neural mechanisms, and the use of phenomenology to assist in identifying the phenotype for genetic studies.
Two different sets of diagnostic criteria are currently used in the world literature. For most studies that emphasize biological markers, and for almost all of those conducted in the United States, the standard diagnostic criteria are from the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV, American Psychiatric Association, 1994). However, international epidemiologic studies are likely to use the World Health Organization's International Classification of Diseases, Tenth Revision (WHO's ICD-10, World Health Organization, 1992). Differences in the choice of diagnostic criteria may affect the results of studies.
There are many similarities between the ICD and DSM, largely as a consequence of efforts by the ICD and DSM work groups to achieve as much concordance as possible. Both require 1 month of active symptoms and the presence of psychotic symptoms such as delusions or hallucinations. There are, however, important differences between the ICD and DSM. In most respects, the DSM provides a slightly narrower conceptualization of schizophrenia than does the ICD. For example, the ICD only requires 1 month of overall duration of symptoms, whereas the DSM requires 6 months. In addition, the ICD includes schizotypal disorder and simple schizophrenia within its nomenclature under the general heading of the diagnosis of schizophrenia. In the DSM, simple schizophrenia is excluded, and schizotypal disorder is placed among the personality disorders. Other less significant differences include a greater emphasis on first-rank symptoms in the ICD, as well as a much more specific and complex system list.
How important is the choice of diagnostic criteria for research on adolescents? It could be very important. Setting criteria boundaries more broadly or more narrowly will have a significant impact on the groups of adolescents chosen for study. Furthermore, although the developers of these criteria paid close attention to examining the reliability and, when possible, their validity, it was assumed almost without question that the criteria could and should be the same for children, adolescents, and adults. This decision was not based on any published empirical data but rather, primarily on “clinical impressions.”
A frequently expressed clinical impression among those who study schizophrenia or psy chosis in children and adolescents, however, is that making a diagnosis in these younger age ranges is much more difficult than diagnosing individuals in their 20s. Multiple issues arise for a diagnosis in the adolescence age range. One important issue is comorbidity. Teenagers frequently may meet criteria for multiple diagnoses, such as conduct disorder or attention-deficit hyperactivity disorder (ADHD). Although the DSM tends to encourage the use of multiple diagnoses, this policy also has no empirical basis. An alternative approach that might be considered from research on adolescents is to try to identify a single “best” diagnosis that would summarize the child parsimoniously.
Many adolescents also abuse substances of many different kinds. This factor is important to consider in the diagnosis of adults who may have schizophrenia, but it poses even greater problems among adolescents. Abuse of substances such as amphetamines may potentially induce a psychotic picture that is very similar to schizophrenia. We do not know whether young people who continue to meet criteria for schizophrenia after discontinuing amphetamine use should be considered “typical schizophrenics” or whether they should in fact be given another diagnosis such as substance-induced psychotic disorder. However, because amphetamines have a significant effect on the dopamine system—a key neurotransmitter implicated in the neurochemical mechanisms of schizophrenia—it is at least plausible that amphetamines (and perhaps other substances as well) be considered triggers or inducers of schizophrenia. According to this view, substance abuse could be one of the many factors that rank among the nongenetic causes of schizophrenia. However, there is still no strong consensus on this issue.
In summary, there are many unanswered research questions under the heading of diagnosis. More studies are needed to explore how well existing diagnostic criteria actually work in children and adolescents. Specifically, studies of both the reliability and validity of these criteria are needed, as well as studies examining issues of comorbidity and longitudinal studies examining changes in both diagnosis and phenomenology in cohorts of adolescents and of adults.
The concept of phenomenology can be relatively broad, describing clinical symptoms, psychosocial functioning, cognitive functioning, and “neurological” measures such as soft signs. Here we will focus primarily on clinical symptoms and psychosocial functioning.
For the assessment of clinical symptoms in schizophrenia, choosing the appropriate informant is a key issue. Whatever the age, patients suffering from schizophrenia frequently have difficulty in reporting their symptoms and past history accurately. Optimally, one gets the best information from several informants, usually a parent plus the patient. In the case of adolescents, a friend may be a good additional informant. Another critical issue in phenomenology when assessing adolescents is to determine the distinction between “normal” adolescent behavior and psychopathology. Again, this can be difficult in assessing adults, but it is even more difficult in adolescents. It can be hard to draw the line between “teenage scruffiness” and disorganization, or a withdrawal to seek privacy versus avolition. As discussed above, drug use or abuse can also confound the picture. For example, when an adolescent known to be using marijuana regularly exhibits chronic apathy and avolition, is this due to marijuana use or is it a true negative symptom? At the moment, no data are available to help us address any of these issues pertaining to the assessment of clinical symptoms in adolescents versus those in adults. This is clearly an area in which more information is needed.
Another issue is the identification of appropriate developmental milestones and needs that are appropriate to the adolescent age range for the assessment of psychosocial functioning. For example, when we assess peer relationships in young or older adults with schizophrenia, we are evaluating the extent to which they have a circle of friends with whom they get together socially. In the case of adolescents, peer relationships are far more important and are more intensely driven by a need to establish independence from the family setting and to bond with others from the same age range. Likewise, the assessment of family relationships among adolescents is guided by quite different conditions than those for mature adults. Finally, the “work” of an adolescent is quite generally to do well in school, whereas the “work” of an adult is normally to find a paying job. Again, assessment tools have simply not been defined for assessing these aspects of psychosocial functioning in adolescents.
Expression of Early Symptoms and Illness Course
A wide range of symptoms has been described (Table 5.1) and the initial clinical features vary from one patient to another. The identification of these as prodromal symptoms is essentially retrospective, being diagnosed only after the first psychotic episode heralds. Using detailed assessment of such symptoms by a structured interview, studies have shown that the prodromal symptoms may begin 2 to 6 years before psychosis onset (Hafner et al., 1992). Negative symptoms of the prodrome may begin earlier than the positive symptoms (Häfner, Maurer, Löffler, & Riecher-Rossler, 1993). Over the past decade, attempts have been made to characterize the prodromal phase prospectively, with operational criteria (Yung & McGorry, 1996). However, several such patients may not develop schizophrenia, leading to the problem of false positives; it is therefore critical that we identify more specific predictors of conversion to psychosis among prospectively identified prodromal patients.
Table 5.1 Prodromal Features in First-Episode Psychosis Frequently Described in Adolescent Patients
Reduced concentration, attention
Decreased motivation, drive, and energy
Mood changes: depression, anxiety
Decline in role functioning, e.g., giving less to academic performance, quitting established interests, neglecting appearance
The onset of the first episode of psychosis (the beginning of clearly evident psychotic symptoms) is to be distinguished from the illness onset, which often begins with symptoms and signs of nonspecific psychological disorder (Häfner et al., 1993). The prodromal phase refers to the period characterized by symptoms marking a change from the premorbid state to the time frank psychosis begins (Fig. 5.1). The onset of both the prodrome and the psychotic episode are difficult to define precisely.
Although the clinical features of adolescent-onset and adult-onset schizophrenia are overall quite similar, early onset of schizophrenia may have an impact on its initial clinical presentation. In general, early-onset schizophrenia patients have more severe negative symptoms and cognitive impairments and are less responsive to treatment. Children and adolescents with schizophrenia often tend to fail in achieving expected levels of academic and interpersonal achievement. Very early-onset cases tend to have an insidious onset, whereas adolescent-onset cases tend to have a more acute onset. Patients with EOS or VEOS are also more often diagnosed as an undifferentiated subtype, because well-formed delusions and hallucinations are less frequent (Nicolson & Rapoport, 1999; Werry, McClellan, & Chard, 1991).
In summary, a challenge in the study of schizophrenia is the variability, or heterogeneity, in the clinical manifestations, and associated biological changes and course. This heterogeneity may have lead to inconsistencies in research findings (Keshavan & Schooler, 1992). Identifying early symptoms and signs and functional impairment can help our efforts in improving early diagnosis and in understanding the biological and genetic heterogeneity. Knowledge of the illness onset in adolescence may also help elucidate the brain developmental and possibly neurodegenerative processes in this illness, as proposed by recent pathophysiological models. Furthermore, an understanding of the course of clinical and neurobiological characteristics in the early phase of schizophrenia, such as the duration of untreated illness, can help in predicting outcome and presents important opportunities for secondary prevention.
Some of the key research questions in the area of the phenomenology of adolescent schizophrenia are as follows:
• Should the same criteria be used for adolescents and adults?
• Are there differences in phenomenology between the two?
• What is the validity of current assessment tools in the “real” research world? The reliability?
• What is (are) the best source(s) of information?
• What impact do differences in “life developmental stages” have on phenomenology?
• What is the best way to assess comorbidity and boundary issues in relationship to other disorders such as schizotypal disorder?
The consensus on these questions is at best modest. Almost no empirical data are available to answer them. In the area of phenomenology in adolescents suffering from schizophrenia, more well-designed, empirical studies are needed to improve assessment tools and to compare adults and adolescents.
Linking Phenomenology to Its Neural Basis
Through the use of neuroimaging, neuropathology, and neurogenetics, substantial progress in understanding the neural underpinnings of schizophrenia is being made. Excellent work has been done recently that examines the relationship between brain development and the occurrence of schizophrenia in children and adolescents, as described in other chapters (DeLisi, 1997; Giedd et al., 1999; Gur, Maany, et al., 1998; 1999; Ho et al., 2003; Jacobsen et al., 1998; Kumra et al., 2000; Lieberman, Chakos, et al., 2001; Rapoport et al., 1997; Thompson, Vidal, et al., 2001). As this work continues to mature, however, more work needs to be done to examine precisely how the specific symptoms of schizophrenia arise in the human brain, and whether imaging and other tools can be used to assist in diagnosis, treatment planning, and ultimately prevention.
This work must also address several questions in the realm of phenomenology. Specifically, how should we proceed as we attempt to link phenomenology to neural mechanisms? As discussed above, the phenomenology has multiple levels and aspects—symptoms, outcome, cognitive function, and psychosocial function. Which of these should be linked to imaging and other “biological” measures?
Most work to date has taken several different approaches. At the simplest level, investigators have conducted studies linking specific symptoms to neural measures. For example, studies have used positron emission tomography (PET) to identify brain regions active during auditory hallucinations (e.g., Silbersweig et al., 1995). Other investigators have examined symptoms such as thought disorder in relation to brain measures (e.g., Shenton et al., 1992). One of the critical conceptual issues, however, is the fact that the phenomenology of schizophrenia is complex. That is, the illness cannot be characterized on the basis of a single symptom. Although auditory hallucinations are common in schizophrenia, they are not omnipresent. Therefore, other investigators have proceeded by examining groups of symptoms that are correlated with one another, or “dimensions.” Many factor analytic studies have examined the factor structure of the symptoms of schizophrenia; nearly all find that the symptoms group naturally into three dimensions: psychoticism, disorganization, and negative symptoms (Andreasen, 1986; Andreasen, O'Leary, et al., 1995; Andreasen, Olsen, & Dennert, 1982; Arndt, Alliger, & Andreasen, 1991; Arndt, Andreasen, Flaum, Miller, & Nopoulos, 1995; Bilder, Mukherjee, Rieder, & Pandurangi, 1985; Gur et al., 1991; Kulhara, Kota, & Joseph, 1986; Lenzenweger, Dworkin, & Wethington, 1989; Liddle, 1987). Some studies have used the dimensional approach to examine brain–behavior relationships. Several studies also suggest that these three dimensions may have different functional neural substrates as seen with PET, or different structural brain correlates as evaluated with magnetic resonance imaging (MRI), and may also have different and independent longitudinal courses (Andreasen, Arndt, Alliger, Miller, & Flaum, 1995; Andreasen et al., 1996, 1997; Arndt et al., 1995; Flaum et al., 1995, 1997; Gur et al., 1991; Miller, Arndt, & Andreasen, 1993; O'Leary et al., 2000).
In concert with this work examining the symptoms of schizophrenia, other investigators have pursued the study of relationships between cognition and brain measures. Some have argued that some form of cognitive dysfunction may ultimately provide the best definition of the phenotype of schizophrenia, and that ultimately cognitive measures may replace symptom measures in defining the phenomenology of schizophrenia (Andreasen, 1999). Again, however, a consensus has not been achieved.
Defining the Phenotype for Genetic Studies
Contemporary geneticists applying the tools of modern genetics have become very much aware of how important it is to have good definitions of complex disorders such as schizophrenia. In fact, reflecting this awareness, they are beginning to speak about a new (but actually old) field, referred to as “phenomics,” the genetic underpinnings of phenomenology. The emergence of this term reflects the fact that the definition of the phenotype of illnesses like schizophrenia may be the single most important component of modern genetic studies.
Here the issues are very similar to those discussed above, involving the relationship between clinical presentation and neural mechanisms. At what level should the phenotype be defined? The symptom level? Dimension level? Diagnosis level? Cognitive level? Or should we abandon these more superficial clinical measurements and attempt to find more basic definitions, often referred to as “endophenotypes,” or “measurable components unseen by the unaided eye along the pathway between disease and distal genotype” (Gottesman & Gould, 2003)?
In this instance, there may be some consensus. Many investigators believe that endophenotypic definitions may provide a better index of the presence of this disorder than classic symptom-based definitions, such as those created by the DSM or ICD. There is as of yet, however, no strong consensus on what the “best” endophenotypes may be. Some candidates that have been proposed include problems with working memory, eye tracking, or prepulse inhibition. To date, most of this work has been conducted with adults. The application of this approach to defining and identifying the schizophrenia endophenotype in children and adolescents is another important future direction, as is the search for additional new candidate endophenotypes.
Two complementary approaches have emerged as providing much needed insight into the causes and underlying substrates of schizophrenia: neurobiology and genetics. Current efforts in neurobiology are to integrate data from behavioral measurements with the increasingly informative data from work with neuroimaging and electrophysiology. Neurobiological studies were stimulated by the well-documented neurobehavioral deficits that are present in schizophrenia. Some of the impairments are evident at the premorbid phase of illness and progress during adolescence, with onset of symptoms. These have become targets for therapeutic interventions. The application of structural and functional neuroimaging has enabled researchers to obtain in vivo measures and highlight the brain circuitry affected in schizophrenia. Progress in genetics has moved the field from earlier efforts relying on family studies of the phenotype to molecular studies that probe the underlying biology. In this section, we will review neurobehavioral measures, proceed to describe studies of brain structure and function, review the impact of hormones critical during adolescence, describe the implicated brain circuitry, and conclude by presenting the genetics of schizophrenia.
Cognitive deficits have been recognized since early descriptions of schizophrenia, when it was called “dementia praecox.” More recent evidence confirms that cognitive deficits are evident in vulnerable individuals, are present at the onset of illness, and predict outcome. Furthermore, as summarized in Chapters 6 and 7, early detection and efforts at intervention may hold a key for ameliorating the ravages of schizophrenia later in life. Here we will describe evidence for deficits in neuromotor and neurocognitive functioning, with special emphasis on early presentation.
Prior to the advent of antipsychotic medications, there were reports in the scientific literature on the occurrence of movement abnormalities in patients with schizophrenia (Huston & Shakow, 1946; Walker, 1994; Yarden & Discipio, 1971). After treatment of patients with antipsychotics became widespread, attention shifted to drug-induced abnormalities in motor behavior. Because motor side effects were of such great concern, they temporarily eclipsed research on naturally occurring motor dysfunction in schizophrenia. But in recent decades, the findings from prospective and retrospective studies have rekindled interest in the signs of motor dysfunction that often accompany schizophrenia in the absence of treatment.
Because the association between motor deficits and brain dysfunction is so well established, motor behaviors are particularly interesting to researchers in the field of schizophrenia (Walker, 1994). In clinical practice, neurologists are often able to identify the locus of brain lesions based on the nature of motor impairments. To date, the motor signs observed in schizophrenia have generally been too subtle and nonspecific to suggest a lesion in a particular brain structure. Nonetheless, there is extensive evidence that motor dysfunction is common in schizophrenia, and it may offer clues about the nature of the brain dysfunction subserving the disorder.
Research has shown that motor deficits predate the onset of schizophrenia, and for some patients are present early in life. Infants who later develop schizophrenia show delays and abnormalities in motor development (Fish, Marcus, Hans, & Auerbach, 1993; Walker, Savoie, & Davis, 1994). They are slower to acquire coordinated patterns of crawling, walking, and bimanual manipulation. They also manifest asymmetries and abnormalities in their movements. These include abnormal postures and involuntary movements of the hands and arms. It is important to note, however, that these early motor signs are not specific to schizophrenia. Delays and anomalies in motor development are present in children who later manifest a variety of disorders, as well as some who show no subsequent disorder. Thus, we cannot use motor signs as a basis for early diagnosis or prediction. But the presence of motor deficits in infants who subsequently manifest schizophrenia suggests that the vulnerability to the disorder involves the central nervous system and is present at birth.
Deficits in motor function extend beyond infancy and have been detected throughout the premorbid period in schizophrenia, including adolescence. Studies of the school and medical records of individuals diagnosed with schizophrenia in late adolescence or early adulthood reveal an elevated rate of motor problems. Both school-aged children and adolescents at risk are more likely to have problems with motor coordination (Cannon, Jones, Huttunen, Tanskanen, & Murray, 1999). Similarly, prospective research has shown that children and adolescents who later develop schizophrenia score below normal controls on standardized tests of motor proficiency (Marcus, Hans, Auerbach, & Auerbach, 1993; Niemi, Suvisaari, Tuulio-Henriksson, & Loennqvist, 2003; Schreiber, Stolz-Born, Heinrich, & Kornhuber, 1992). Again, the presence of these deficits before the onset of clinical schizophrenia suggests that they are indicators of biological vulnerability.
As mentioned, there is an extensive body of research on motor functions in adult patients diagnosed with schizophrenia, both medicated and nonmedicated (Manschreck, Maher, Rucklos, & Vereen, 1982; Walker, 1994; Wolff & O'Driscoll, 1999). The research has revealed deficits in a wide range of measures, from simple finger tapping to the execution of complex manual tasks. In addition, when compared to healthy comparison subjects, schizophrenia patients manifest more involuntary movements and postural abnormalities.
It is noteworthy that motor abnormalities have also been detected in adolescents with schizotypal personality disorder. Compared to healthy adolescents, these children show more involuntary movements and coordination problems (Nagy & Szatmari, 1986; Walker, Lewis, Loewy, & Palyo, 1999). Further research is needed to determine whether schizotypal adolescents with motor abnormalities are more likely to succumb to schizophrenia.
The nature of the motor deficits observed in schizophrenia suggests abnormalities in subcortical brain areas, in particular a group of brain regions referred to as the basal ganglia (Walker, 1994). These brain regions are a part of the neural circuitry that connects subcortical with higher cortical areas of the brain. It is now known that the basal ganglia play a role in cognitive and emotional processes, as well as motor functions. As our understanding of brain function and motor circuitry expands, we will have greater opportunities for identifying the origins of motor dysfunction in schizophrenia. In addition, research on motor abnormalities in schizophrenia has the potential to shed light on the neural substrates that confer risk for schizophrenia. Some of the important questions that remain to be answered are: What is the nature and prevalence of motor dysfunction in adolescents at risk for schizophrenia? Is the presence of motor dysfunction in schizophrenia linked with a particular pattern of neurochemical or brain abnormalities? Can the presence of motor dysfunction aid in predicting which individuals with prodromal syndromes, such as schizotypal personality disorder (SPD), will develop schizophrenia? Would neuromotor assessment aid in the prediction of treatment response?
Early studies examining cognitive function in schizophrenia focused on single domains, such as attention or memory, and preceded developments in neuroimaging and cognitive neuroscience that afford better linkage between cognitive aberrations and brain circuitry. Neuropsychological batteries, which have been initially developed and applied in neurological populations, attempt to link behavioral deficits to brain function. When applied in schizophrenia, such batteries have consistently indicated diffuse dysfunction, with relatively greater impairment in executive functions and in learning and memory (Bilder et al., 2000; Censits, Ragland, Gur, & Gur, 1997; Elvevag & Goldberg, 2000; Green, 1996; Gur et al., 2001; Saykin et al., 1994).
It is noteworthy that the pattern of deficits is already observed at first presentation and is not significantly changed by treatment of the clinical symptoms. Therefore, study of adolescents at risk or at onset of illness avoids confounding by effects of treatment, hospitalization, and social isolation that may contribute to compromised function. Although the literature evaluating the specificity of cognitive deficits in schizophrenia is limited, there is enough evidence to show that the profile and severity are different from bipolar disorder. Thus, early evaluation during adolescence may have diagnostic and treatment implications. Given the evidence on cognitive deficits at the premorbid stage, it would be important to evaluate whether a pattern of deficits in adoles cents at risk can predict the onset and course of illness. The executive functions impaired in adults with schizophrenia are the very abilities that are essential for an adolescent to make the transition to young adulthood, when navigation through an increasing complexity of alternatives becomes the issue.
In addition to the cognitive impairment, emotion-processing deficits in identification, discrimination, and recognition of facial expressions have been observed in schizophrenia (Kohler et al., 2003; Kring, Barrett, & Gard, 2003). Such deficits may contribute to the poor social adjustment already salient before disease onset. Emotional impairment in schizophrenia is clinically well established, manifesting in flat, blunted, inappropriate affect and in depression. These affect-related symptoms are notable in adolescents during the prodromal phase of illness preceding the positive symptoms. While these may represent a component of the generalized cognitive impairment, they relate to symptoms and neurobiological measures that deserve further research.
Several brain systems are implicated by these deficits. The attention-processing circuitry includes brainstem-thalamo-striato-accumbens-temporal-hippocampal-prefrontal-parietal regions. Deficits in working memory implicate the dorsolateral prefrontal cortex, and the ventromedial temporal lobe is implicated by deficits in episodic memory. A dorsolateral-medial-orbital prefrontal cortical circuit mediates executive functions. Animal and human investigations have implicated the limbic system, primarily the amygdala, hypothalamus, mesocorticolimbic dopaminergic systems, and cortical regions including orbitofrontal, dorsolateral prefrontal, temporal, and parts of parietal cortex. These are obviously complex systems and impairment in one may interact with dysfunction in others. Studies with large samples are needed to test models of underlying pathophysiology.
The link between neurobehavioral deficits and brain dysfunction can be examined both by correlating individual differences in performance with measures of brain anatomy and through the application of neurobehavioral probes in functional imaging studies. With these paradigms, we can investigate the topography of brain activity in response to engagement in tasks in which deficits have been noted in patients. Thus, there is “online” correlation between brain activity and performance in a way that permits direct examination of brain–behavior relations (Gur et al., 1997).
The availability of methods for quantitative structural neuroimaging has enabled examination of neuroanatomic abnormalities in schizophrenia. Because the onset of schizophrenia takes place during a phase of neurodevelopment characterized by dynamic and extensive changes in brain anatomy, establishment of the growth chart is necessary to interpret findings. Two complementary lines of investigation have proved helpful. By examining the neuroanatomical differences between healthy people and individuals with childhood-onset and first-episode schizophrenia, as well as individuals at risk, regional abnormalities early in the course of illness may be identified. Complementary efforts are needed to examine changes associated with illness progression. An understanding of the neuroanatomic changes in the context of the dynamic transitions of the developing brain during adolescence, however, requires careful longitudinal studies during this critical period. A brief introduction to the methodology of quantitative MRI and its application to examine neurodevelopment is needed to appreciate findings in schizophrenia.
Several approaches have been developed in the early 1990s, and these have now become standard and have been shown to produce reliable results (e.g., Filipek, Richelme, Kennedy, & Caviness, 1994; Kohn et al., 1991). These methods have provided data on the intracranial composition of the three main brain compartments related to cytoarchitecture and connectivity: gray matter (GM), the somatodendritic tissue of neurons (cortical and deep); white matter (WM), the axonal compartment of myelinated connecting fibers; and cerebrospinal fluid (CSF).
In one of the first studies examining segmented MRI in children and adults, Jernigan and Tallal (1990) documented the “pruning” process proposed by Huttenlocher's (1984) work. They found that children had higher GM volumes than adults, a finding indicating loss of GM during adolescence. This group has more recently replicated these results by use of advanced methods for image analysis (Sowell, Thompson, Holmes, Jernigan, & Toga, 1999). Their new study also demonstrated that the pruning is most “aggressive” in prefrontal and temporoparietal cortical brain regions. As a result of this work, we now recognize that both myelination and pruning are important aspects of brain development.
In a landmark paper published in 1996, a National Institutes of Health (NIH) group reported results of a brain volumetric MRI study on 104 healthy children ranging in age from 4 to 18 (Giedd et al., 1996). Although this group did not segment the MRI data into compartments, they did observe developmental changes that clearly indicated prolonged maturation beyond age 17. In a later report on this sample, in which segmentation algorithms were applied, the investigators were able to pinpoint the greatest delay in myelination, defined as WM volume, for frontotemporal pathways (Paus et al., 1999). This finding is very consistent with the Yakovlev and Lecours (1967) projections. The NIH group went on to exploit the ability of MRI to obtain repeated measures on the same individuals. Using these longitudinal data, they were able to better pinpoint the timing of preadolescent increase in GM that precipitates the pruning process of adolescence. Of importance to the question of maturation as defined by myelogenesis are results indicating that the volume of WM continued to show increases up to age 22 years (Giedd et al., 1999).
A Harvard group developed a sophisticated procedure for MRI analysis (Filipek et al., 1994) which they applied to a sample of children with the age range of 7 to 11 years and used to compare results with those of adults (Caviness, Kennedy, Richelme, Rademacher, & Filipek, 1996). They found sex differences suggesting earlier maturation of females, and generally supported the role of WM as an index of maturation. Their results also indicated that WM shows a delay in reaching its peak volume until early adulthood.
Another landmark study, published by a Stanford group, examined segmented MRI on a “retrospective” sample of 88 participants ranging in age from 3 months to 30 years and a “prospective” sample of 73 healthy men aged 21 to 70 years (Pfefferbaum et al., 1994). Scans for the retrospective sample were available from the clinical caseload, although images were carefully selected to include only those with a negative clinical reading; the prospective sample was recruited specifically for research and was medically screened to be healthy. The results demonstrated a clear neurodevelopmental course for GM and WM, the former showing a steady decline during adolescence whereas the latter showed increased volume until about age 20 to 22 years.
A Johns Hopkins group used a similar approach in a sample of 85 healthy children and adolescents ranging in age from 5 to 17 years (Reiss, Abrams, Singer, Ross, & Denckla, 1996). Consistent with postmortem and the other volumetric MRI studies, these investigators reported a steady increase in WM volume with age that did not seem to peak by age 17. Unfortunately, they did not have data on older individuals. Their results are consistent with those of Blatter et al. (1995) from Utah, although the extensive Utah database combines ages 16 to 25 and therefore does not permit evaluation of changes during late adolescence and early adulthood.
In the only study to date that has examined segmented MRI volumes from a prospective sample of 28 healthy children aged 1 month to 10 years and a small adult sample, Matsuzawa et al. (2001) applied the segmentation procedures developed by the Penn group. Matsuzawa et al. demonstrated increased volume of both GM and WM in the first postnatal months, but whereas GM volume peaked at about 2 years of age, the volume of WM, which indicates brain maturation, continued to increase into adulthood (Figure 5.2). Furthermore, consistent with the postmortem and other MRI studies that have examined this issue, the frontal lobe showed the greatest maturational lag, and its myelination is unlikely completed before young adulthood.
Magnetic resonance imaging studies in first-episode patients have indicated smaller brain volume and an increase in CSF relative to that in healthy people (e.g., Gur et al. 1998a; Ho et al., 2003). The increase is more pronounced in ventricular than in sulcal CSF. Brain and CSF volumes have been related to phenomenological and other clinical variables such as premorbid functioning, symptom severity, and outcome. Abnormalities in these measures are likely to be more pronounced in patients with poorer premorbid functioning, more severe symptoms, and worse outcome. The concept of brain reserve or resilience may apply to schizophrenia as well, with normal brain and CSF volumes as preliminary indicators of protective capacity. As our understanding of how brain systems regulate behavior in health and disease improves, we can take advantage of neuroimaging to examine specific brain regions implicated in the pathophysiology of schizophrenia.
Gray and white matter tissue segmentation can help determine whether tissue loss and disorganization in schizophrenia are primarily the result of a GM deficit or whether abnormalities in WM are also involved. Several studies using segmentation methods have indicated that GM volume reduction characterizes individuals with schizophrenia, whereas the volume of WM is normal. The reduction in GM is apparent in first-episode, never-treated patients and supports the growing body of work that schizophrenia is a neurodevelopmental disorder (e.g., Gur et al., 1999).
In evaluating specific regions, the most consistent findings are of reduced volumes of prefrontal cortex and temporal lobe structures. Other brain regions also noted to have reduced volumes include the parietal lobe, thalamus, basal ganglia, cerebellar vermis, and olfactory bulbs. Relatively few studies have related sublobar volumes to clinical or neurocognitive measures. Available studies, however, support the hypothesis that increased volume is associated with lower severity of negative symptoms and better cognitive performance (e.g., Gur et al., 2000a,b; Ho et al., 2003).
The question of progression of tissue loss has been addressed in relatively few studies and in small samples, reflecting the difficulty of recruiting for study patients in the early stages of illness. Longitudinal studies applying MRI have examined first-episode patients. One group of investigators found no ventricular changes in a follow-up study, conducted 1 to 2 years after the initial study, of 13 patients and 8 controls (De greef et al., 1992). Another study evaluated 16 patients and 5 controls, studied 2 years after a first psychotic episode (DeLisi et al., 1991). Patients showed no consistent change in ventricular size with time, although there were individual increases or decreases. With a slightly larger group of 24 patients and 6 controls, no significant changes were observed in ventricular or temporal lobe volume at follow-up (DeLisi et al., 1992). Subsequently, 20 of these patients and 5 controls were rescanned over 4 years, and greater decreases in whole-brain volume and enlargement in left ventricular volume were observed in patients. The authors concluded that subtle cortical changes may occur after the onset of illness, suggesting progression in some cases (DeLisi et al., 1995).
In a longitudinal study with a larger sample, 40 patients (20 first-episode, 20 previously treated) and 17 healthy participants were rescanned an average of 2.5 years later. Volumes of whole brain, CSF, and frontal and temporal lobes were measured (Gur, Cowell, et al., 1998). First-episode and previously treated patients had smaller whole-brain, frontal, and temporal lobe volumes than controls at intake. Longitudinally, a reduction in frontal lobe volume was found only in patients, and was most pronounced at the early stages of illness, whereas temporal lobe reduction was seen also in controls. In both first-episode and previously treated patients, volume reduction was associated with decline in some neurobehavioral functions.
The question of specificity of neuroanatomic findings to schizophrenia was addressed in a recent study that evaluated 13 patients with first-episode schizophrenia, 15 patients with first-episode affective psychosis (mainly manic), and 14 healthy comparison subjects longitudinally, with scans separated by 1.5 years (Kasai et. al., 2003a). The investigators reported that patients with schizophrenia had progressive decreases in GM volume over time in the left superior temporal gyrus, compared with that in both of the other groups. The existence of neuroanatomical abnormalities in first-episode patients indicates that brain dysfunction occurs before clinical presentation. However, the longitudinal studies suggest evidence of progression, in which anatomic changes may impact some clinical and neurobehavioral features of the illness in some patients. There is also evidence that progression is significantly greater in early-onset patients during adolescence than it is for adult subjects (Gogate, Giedd, Janson, & Rapoport, 2001).
Findings from MRI have been most consistent for GM volume reduction, but more recently, WM changes have also been reported. In the coming years the availability of diffusion tensor imaging will enhance the efforts to examine compartmental abnormalities. The growing understanding of brain development and MRI data obtained from children suggest that the neuroanatomic neuroimaging literature in schizophrenia is consistent with diffuse disruption of normal maturation. Thus, there is clear evidence for structural abnormalities in schizophrenia that are associated with reduced cognitive capacity and less clearly with symptoms. Future work, perhaps with more advanced computerized parcellation methods, is needed to better chart the brain pathways most severely affected.
The electroencephalogram (EEG) measures the electrical activity of the brain; it originates from the summated electrical potentials generated by inhibitory and excitatory inputs onto neurons. The main source of the scalp-recorded EEG is in the cortex of the brain, which contains the large and parallel dendritic trees of pyramidal neurons whose regular ordering facilitates summation. One of the important advances in EEG-based research was the development of a technique to isolate the brain activity related to specific events from the background EEG; this activity related to specific events is termed event-related potentials, or ERPs. Using averaging techniques, it is possible to visualize events related to one of the many different brain operations reflected in the EEG. Typically, these ERPs are related to the specific processing of certain sensory stimuli.
In recent years, many new means of measuring brain structure and function have been developed, each with its advantages in study of the brain. Electroencephalographic and ERP measures are unsurpassed in providing real-time, millisecond resolution of normal and pathological brain processing, literally at the speed of thought, whereas functional magnetic resonance imaging (fMRI) and PET have temporal resolutions some thousand-fold less. Moreover, fMRI and PET only indirectly track neural activity through its effects on blood flow or metabolism. However, the ability of EEG and ERP techniques to localize sources of activity is much less than that of fMRI and PET, and these methods, together with structural MRI, are needed to supplement EEG and ERP information.
Current Event-Related Potential Research in Schizophrenia
Space limitations preclude discussion of all ERPs. We provide here a sample of current work designed to illuminate a fundamental question in schizophrenia research—namely, how the brains of patients suffering from this disorder differ from those of healthy subjects. Event-related potentials provide a functional window on many aspects of brain processing. These include the most elementary ones, likely involving cellular circuitry (gamma band activity), early, simple signal detection and gating (P50), and automatic detection of changes in the environment (mismatch negativity activity), and more complex activity such as conscious updating of expectations in view of unusual events (P300).
In this section we will first briefly review studies of ERP processes in adults with schizophrenia that illustrate the potential of these measures to provide clues about the cellular circuitry that may be impaired in schizophrenia. The auditory modality plays a special role because it is severely affected in schizophrenia, as evinced in the primacy of auditory hallucinations and speech and language pathology. The data presented here support the hypothesis that schizophrenia involves abnormalities in brain processing from the most simple to the most complex level, and that the anatomical substrates of auditory processing in the neocortical temporal lobe, most carefully investigated in the superior temporal gyrus, themselves evince reduction in GM volume. Next, we briefly summarize a series of studies of adolescents with schizophrenia in which ERPS are recorded while the youngsters perform poorly on cognitive tasks that make extensive demands on processing resources. These studies use ERPs in an attempt to identify the earliest stage of cognitive processing at which deficits emerge in adolescents with schizophrenia.
Gamma-band activity and neural circuit abnormalities at the cellular level.
The first ERP we will consider is the steady-state gamma-band response. Gamma band refers to a brain oscillation at and near the frequency of 40 Hertz (Hz) or 40 times per second; steady-state refers to its being elicited by a stimulus of the same frequency. At the cellular level, gamma-band activity is an endogenous brain oscillation thought to reflect the synchronizing of activity in several columns of cortical neurons, or between cortex and thalamus, with this synchronization facilitating communication. At the cognitive level, work in humans suggests that gamma activity reflects the convergence of multiple processing streams in cortex, giving rise to a unified percept. A simple example is a “fire truck”; a particular combination of form perception, motion perception, and auditory perception are melded to form this percept. Gamma activity at its simplest, however, involves basic neural circuitry composed of projection neurons, usually using excitatory amino acid (EAA) neurotransmission, linked with inhibitory gamma-aminobutyric acid (GABA)ergic interneurons. Studies of gamma activity in schizophrenia aim to determine if there is a basic circuit abnormality present, such as might arise from a deficiency in recurrent inhibition, postulated by a number of workers (see review in McCarley, Hsiao, Freedman, Pfefferbaum, & Donchin, 1996). Gamma-band studies themselves, however, cannot reveal any specific details of neural circuitry abnormality.
Kwon and colleagues (1999) began the study of gamma in schizophrenia using an exogenous input of 40-Hz auditory clicks, leading to a steady-state gamma response. The magnitude of the brain response was measured by power, the amount of EEG energy at a specific frequency, with the degree of capability of gamma driving being reflected in the power at and near 40 Hz. Compared with healthy controls, schizophrenia patients had a markedly reduced power at 40-Hz input, although they showed normal driving at slower frequencies, which indicated that this was not a general reduction in power but one specific to the gamma band.
Spencer and colleagues (2003) took the next logical step and evaluated the gamma-band response to visual stimuli in schizophrenia, to determine whether high-frequency neural synchronization associated with the perception of visual gestalts is abnormal in schizophrenia patients. Previous studies of healthy individuals had reported enhancements of gamma-band power (Tallon-Baudry & Bertrand, 1999) and phase locking (Rodriguez et al., 1999) when gestalt objects are perceived. In the study by Spencer et al., individuals with schizophrenia and matched healthy people discriminated between square gestalt stimuli and non-square stimuli (square/no-square conditions). In schizophrenia patients, the early visual system gamma-band response to gestalt square stimuli was lacking. There were also abnormalities in gamma-band synchrony between brain regions, with schizophrenia patients showing decreasing rather than increasing gamma-band coherence between posterior visual regions and other brain regions after perceiving the visual gestalt stimuli. These findings support the hypothesis that schizophrenia is associated with a fundamental abnormality in cellular neural circuitry evinced as a failure of gamma-band synchronization, especially in the 40-Hz range.
Sensory gating and the P50—early sensory gating.
Several ERPs have been related to the search for an electrophysiologic concomitant of an early sensory gating deficit in schizophrenia. These include, for example, the startle response, for which the size of a blink to an acoustic probe is measured. Schizophrenia patients appear to be unable to modify their large startle response when forewarned that a probe is coming, in contrast with controls (e.g., Braff et al., 1978).
Another ERP thought to be sensitive to an early sensory gating abnormality in schizophrenia is the P50. In the sensory gating paradigm, an auditory click is presented to a subject, eliciting a positive deflection about 50 msec after stimulus onset, the P50 component. After a brief interval (about 500 msec), a second click elicits a much smaller-amplitude P50 in normal adult subjects, who are said to show normal gating: the first stimulus inhibits, or closes the gate to, neurophysiological processing of the second stimulus. Patients with schizophrenia, by contrast, show less reduction in P50 amplitude to the second click, which is referred to as a failure in gating (Freedman, Adler, Waldo, Oachtman, & Franks, 1983). This gating deficit occurs in about half the first-degree relatives of a schizophrenic patient, a finding suggesting that it may index a genetic factor in schizophrenia in the absence of overt psychotic symptoms (Waldo et al., 1991). Patients with affective disorder may show a gating deficit, but the deficit does not persist after successful treatment; in patients with schizophrenia, the deficit occurs in both medicated and unmedicated patients and persists after symptom remission (Adler et al., 1991; Freedman et al., 1983).
The gating effect is thought to take place in temporal lobe structures, possibly the medial temporal lobe (Adler, Waldo, & Freedman, 1985). P50 gating is enhanced by nicotinic cholinergic mechanisms, and it is possible that smoking in patients with schizophrenia is a form of self-medication. Freedman et al. (1994) have shown that blockade of the α7 -nicotinic receptor, localized to hippocampal neurons, causes loss of the inhibitory gating response to auditory stimuli in an animal model. The failure of inhibitory mechanisms to gate sensory input to higher-order processing might result in “sensory flooding,” which Freedman suggests may underlie many of the symptoms of schizophrenia.
Mismatch negativity and postonset progression of abnormalities.
Mismatch negativity (MMN) is a negative ERP that occurs about 0.2 sec after infrequent sounds (deviants) are presented in the sequence of repetitive sounds (standards). Deviant sounds may differ from the standards in a simple physical characteristic such as pitch, duration, intensity, or spatial location. Mismatch negativity is primarily evoked automatically, that is, without conscious attention. Its main source is thought to be in or near primary auditory cortex (Heschl gyrus) and to reflect the operations of sensory memory, a memory of past stimuli used by the auditory cortex in analysis of temporal patterns.
There is a consistent finding of a reduction in amplitude of MMN in chronically ill schizophrenia patients that appears to be traitlike and not ameliorated by either typical (haloperidol) or atypical (clozapine) medication (Umbricht et al., 1998). A point of particular interest has been the finding that the MMN elicited by tones of different frequency (the pitch MMN) is normal in patients at the time of first hospitalization (Salisbury, Bonner-Jackson, Griggs, Shenton, & McCarley, 2001; confirmed by Umbricht, Javitt, Bates, Kane, & Lieberman, 2002), whereas the MMN elicited by the same stimuli is abnormal in chronic schizophrenia. This finding suggests that pitch MMN might index a postonset progression of brain abnormalities. Indeed, the prospective longitudinal study of Salisbury, McCarley, and colleagues (unpublished data) now has preliminary data showing that schizophrenia subjects without a MMN abnormality at first hospitalization develop an abnormality over the next 1.5 years.
In the same group of patients, the Heschl gyrus, the likely source of the MMN, demonstrates a progressive reduction in GM volume over the same time period (Kasai et al., 2003b). In participants with both MRI and MMN procedures, the degree of GM volume reduction was found to parallel the degree of MMN reduction, although the number of subjects examined is currently relatively small and this conclusion is tentative. Although the presence of postonset progression of abnormalities is controversial in the field, it is of obvious importance to our understanding of the disorder and of particular importance to the study of adolescents with onset of schizophrenia, because it would prompt a search for possible medication and/or psychosocial treatment that might ameliorate progression.
Recent multimodal imaging (Wible et al., 2001) has demonstrated the presence of a deficiency of fMRI activation (BOLD) in schizophrenia to the mismatch stimulus within Heschl's gyrus and nearby posterior superior temporal gyrus.
Because MMN may reflect, in part, N-methyl-d -aspartate (NMDA)-mediated activity, a speculation about the reason for progression is that NMDA-mediated excitotoxity might cause both a reduction in the neuropil (dendritic regression) and a concomitant reduction in the MMN in the months following first hospitalization. Only further work will determine whether this speculation is valid. It is noteworthy that the MMN abnormalities present in schizophrenic psychosis are not present in manic psychosis.
P300 and the failure to process unusual events.
The P300 is an ERP that occurs when a low-probability event is detected and consciously processed. Typically, subjects are asked to count a low-probability tone that is interspersed with a more frequently occurring stimulus. The P300 differs from the typical MMN paradigm in that the stimuli are presented at a slower rate (typically around one per second) and the subject is actively and consciously attending and processing the stimuli, whereas the MMN stimuli are not consciously processed. P300 is larger when the stimulus is rare. Whereas MMN is thought to reflect sensory memory, by definition preconscious, P300 is thought to reflect an updating of the conscious information-processing stream and of expectancy.
Reduction of the P300 amplitude at midline sites is the most frequently replicated abnormality in schizophrenia, although P300 reduction is also found in some other disorders. This widespread P300 reduction also appears to be traitlike and an enduring feature of the disease. For example, Ford and colleagues (1994) demonstrated that although P300 showed moderate amplitude increases with symptom resolution, it did not approach normal values during these periods of remission. Umbricht et al. (1998) have reported that atypical antipsychotic treatment led to a significant increase of P300 amplitudes in patients with schizophrenia.
In addition to the midline P300 reduction, both chronically ill and first-episode schizophrenic subjects display an asymmetry in P300 with smaller voltage over the left temporal lobe than over the right. The more pronounced this left temporal P300 amplitude abnormality, the more pronounced is the extent of psychopathology, as reflected in thought disorder and paranoid delusions (e.g., McCarley et al., 1993, 2002). It is possible the increased delusions reflect a failure of veridical updating of cognitive schemata. This left temporal deficit is not found in affective (manic) psychosis.
There are likely several bilateral brain generators responsible for the P300, with a generator in the superior temporal gyrus (STG) likely under lying the left temporal deficit, since, in schizophrenia, the greater the reduction in GM volume in posterior STG, the greater the reduction in P300 amplitude at left temporal sites in both chronic and first-episode schizophrenia patients. It is of note that the posterior STG, on the left in right-handed individuals, is an area intimately related to language processing and thinking (it includes part of Wernicke's area), and an area where volume reductions are associated with increased thought disorder and severity of auditory hallucinations.
Event-related Potential Measures in Children and Adolescents with Schizophrenia
Event-related indices of information processing deficits.
Brain activity reflected in ERPs recorded during performance of information-processing tasks can be used to help isolate the component or stage of information-processing that is impaired in schizophrenia. A series of ERP studies of children and adolescents with schizophrenia, conducted by the UCLA Childhood Onset Schizophrenia program, are summarized below (see Strandburg et al., 1994a and Asarnow, Brown, & Strandburg, 1995 for reviews). These studies examined ERP components while children and adolescents with schizophrenia performed tasks like the span of apprehension (Span; Strandburg, Marsh, Brown, Asarnow, & Guthrie, 1984) and a continuous performance test (CPT; Strandburg et al., 1990). Several decades of studying mental chronometry with ERPs has produced a lexicon of ERP components with well-established neurocognitive correlates (Hillyard & Kutas, 1983). These ERP components can be used to help identify the stages of information processing that are impaired in schizophrenia.
The UCLA ERP studies have focused primarily on four components: contingent negative variation (CNV), hemispheric asymmetry in the amplitude of the P1/N1 component complex, processing negativity (Np), and a late positive component (P300). The CNV measures orienting, preparation, and readiness to respond to an expected stimulus. There are at least two separate generators of the CNV: an early frontal component believed to be an orienting response to warning stimuli, and a later central component associated with preparedness for stimuli-processing and response (Rohrbaugh et al., 1986).
Healthy individuals typically have larger visual P1/N1 components over the right cerebral hemisphere. Many of the UCLA studies compared hemispheric laterality between healthy and schizophrenia individuals. Differences in lateralization during visual information-processing tasks could reflect either differences in the strategic use of processing capacity of the hemispheres or a lateralized neural deficit.
The Np is a family of negative components that occur within the first 400 msec after the onset of a stimulus, indicating the degree to which attentional and perceptual resources have been allocated to stimulus processing. Because the Np waves occur contemporaneously with other components (P1, N1, and P2), they are best seen in difference potentials resulting from the subtraction of non-attend ERPs from attend ERPs (Hillyard & Hansen, 1986; Naatanan, 1982). Finally, as described above, the P300 is a frequently studied index of the recognition of stimulus significance in relation to task demands.
Event-related potential results in child and adolescent schizophrenia.
Table 5.2 summarizes by component the ERP results from six UCLA studies of children or adults with schizophrenia. In all the studies summarized in this table there were large and robust performance differences between groups in both the accuracy and reaction times of signal detection responses. Thus, the behavioral paradigms were successful in eliciting information-processing deficits in these patients.
Table 5.2 Information-Processing Tasks in Child and Adolescent and Adult-Onset Schizophrenia: Summary of Evoked Potential Studies.
Strandburg et al., 1984
Norm > schiz
Norm > schiz
Norm > schiz
Norm > schiza
Strandburg et al., 1990
Norm > schiz
Norm = schiz
Norm > schiz
Strandburg et al., 1991
Norm = schiz
Norm > schiz
Norm > schiz
Strandburg et al., 1994a
Norm > schiz
Schiz > norm
Norm > schiz
Norm > schiz
Strandburg et al., 1994b
Norm > schizb
Norm = schiz
Norm > schiz
Norm > schiz
Strandburg et al., 1997
Norm > schiz
Norm > schiz
a Larger task-difficulty increased more in N1 amplitude in normals than in schizophrenics.
b Normals had larger P300 than schizophrenics for targets in the single-target CPT task.
CNV, contingent negative variation; CPT, continuous performance task; Np, processing negativity; P300, late positive component.
The CNV differences between normals and schizophrenics were not consistently found across studies. In the span task (which includes a warning interval) all possible results were obtained (normals > schizophrenics; normals = schizophrenics; and normals < schizophrenics). For the CNV-like negative wave occurring in the CPT task, no group differences were found in either experiment. Because the warning interval was short and the wave was largest frontally, the CNVs in both tasks were most likely the early wave related to orienting. Thus, differences in prestimulus orienting do not seem to reliably ac count for the poor performance of schizophrenics on these tasks. There are mixed results in CNV experiments on adults with schizophrenia, although most studies found smaller CNVs in schizophrenics (Pritchard, 1986). A longer warning interval than that used in the UCLA experiments (500 msec in the span and 1250 msec ISI in the CNV) may be required to detect preparatory abnormalities in schizophrenia.
In every study summarized in Table 5.2 in which processing negativities were measured, Nps were found to be smaller in schizophrenics. This deficit was seen in both children and adults, with both the span and CPT (Strandburg et al., 1994c) tasks. In contrast, a group of children with ADHD studied while they performed a CPT task showed no evidence of a smaller Np. Diminished Np amplitude is the earliest consistent ERP index of schizophrenia-related information-processing deficit in the UCLA studies. These results suggest impaired allocation of attentional and perceptual resources.
Most studies of processing negativities during channel selective attention tasks (Nd) find that adults with schizophrenia produce less attentional-related endogenous negative activity than do normal controls (see reviews by Cohen, 1990, and Pritchard, 1986). The UCLA results compliment this finding in adults by using a discriminative processing task and extend these findings to childhood-and adolescent-onset schizophrenia. Reductions in the amplitude of Np in schizophrenia result from im-pairments in executive functions responsible for the maintenance of an attentional trace (Baribeau-Braun, Picton, & Gosselin, 1983; Michie, Fox, Ward, Catts, & McConaghy, 1990). Baribeau-Braun et al. (1983) observed normal Nd activity with rapid stimulus presentation rates, but reduced amplitudes with slower rates, findings suggesting that the neural substrates of Nd are intact but improperly regulated in schizophrenia. Individuals with frontal lobe lesions resemble individuals with schizophrenia in this regard, in that both groups do not show increased Np to attended stimuli in auditory selection tasks (Knight, Hillyard, Woods, & Neville 1981).
As noted earlier, reduced amplitude P300 in schizophrenic adults has been consistently found using a wide variety of experimental paradigms (Pritchard et al., 1986). As can be seen in Table 5.2, the UCLA studies also consistently observed smaller P300 amplitude in studies of both schizophrenic children and adults, in the span, CPT, and idiom recognition tasks. P300 latency was also measured in two of these studies. Although prolonged P300 latency was found in one study (Strandburg et al., 1994c), no differences were found in another (Strandburg et al., 1994b). The majority of ERP studies have reported normal P300 latency in schizophrenics (Pritchard, 1986).
Absence of right-lateralized P1/N1 amplitude in visual ERPs has been a consistent finding in all five of the UCLA studies that used the CPT and span tasks. Abnormally lateralized electrophysiological responses, related either to lateralized dysfunction in schizophrenia or a pathology-related difference in information-processing strategy, is a consistent aspect of both adult-and childhood-onset schizophrenia. These results are consistent with abnormal patterns of hemispheric laterality in schizophrenics (e.g., Tucker & Williamson, 1984).
In summary, ERP studies of schizophrenic adults and children performing discriminative processing tasks suggest that the earliest reliable electrophysiological correlate of impaired discriminative processing in schizophrenia is the Np component. It appears that children and adolescents with schizophrenia are deficient in the allocation of attentional resources necessary for efficient and accurate discriminative processing. Although diminished amplitude processing negativities have been observed in ADHD in auditory paradigms (Loiselle, Stamm, Maitinsky, & Whipple, 1980; Satterfield, Schell, Nicholar, Satterfield, & Freese, 1990), Np was found to be normal in ADHD children during the UCLA CPT task (Strandburg et al., 1994a). Diminished Np visual processing may be specific to schizophrenic pathology. Later ERP abnormalities in schizophrenia (e.g., diminished amplitude P300) may be a “downstream” product of the uncertainty in stimulus recognition created by previous discriminative difficulties, or they may be one of additional neurocognitive deficits. Abnormalities in later ERP components are not specific to schizophrenia, having been reported in studies of ADHD children (reviewed by Klorman, 1991).
The absence of P1/N1 asymmetry in the visual ERPs of schizophrenics is contemporaneous with diminished Np. However, the fact that Np amplitude varies with the processing demands of the task, whereas P1/N1 asymmetry does not, suggests that the Np deficit plays a greater role in the information-processing deficits manifested by children and adolescents with schizophrenia.
Magnetoencephalography—A Complement to Electroencephalography
Magnetoencephalography (MEG) is the measure of magnetic fields generated by the brain. A key difference between the physical source of the MEG and that of the EEG is that the MEG is sensitive to cells that lie tangential to the brain surface and consequently have magnetic fields oriented tangentially. Cells with a radial orientation (perpendicular to the brain surface) do not generate signals detectable with MEG. The EEG and MEG are complementary in that the EEG is most sensitive to radially oriented neurons and fields. This distinction arises, of course, because magnetic fields are generated at right angles to electrical fields. One major advantage that magnetic fields have over electrical potentials is that, once generated, they are relatively invulnerable to intervening variations in the media they traverse (i.e., the skull, gray and white matter, and CSF), unlike electrical fields, which are “smeared” by different electrical conductivities. This has made MEG a favorite technology for use in source localization, in which attention has been especially focused on early potentials.
Perhaps because of the expense and nonmobility of the recording equipment needed for MEG, there has been relatively little work using MEG in schizophrenia to replicate and extend the findings of ERPs. A search of Medline in 2000 revealed only 23 published studies using MEG measures of brain activity in schizophrenia. The extant studies have shown interesting results. Reite and colleagues demonstrated that M100 component (the magnetic analogue to the N100) showed less interhemispheric asymmetry in male schizophrenics and had different source orientations in the left hemisphere. The recent review by Reite, Teale, and Rojas (1999) should be consulted for more details of the work on MEG in schizophrenia.
In summary, electrophysiology has the advantage of providing real-time information on brain processing, with a resolution in the millisecond range. In schizophrenia, it shows abnormalities of processing from the very earliest stages (Np, mismatch negativity, P50, gamma activity) to later stages of attentive discrepancy processing (P300) and semantic processing (N400). This suggests a model of disturbance that encompasses a wide variety of processing and is most compatible with a brain model of circuit abnormalities underlying processing at each stage, particularly in the auditory modality. This is also compatible with MRI studies of abnormal GM regions associated with abnormal ERPs.
One of the more intriguing potential applications to schizophrenia in adolescence is using ERPs to track progression of brain abnormalities. The mismatch negativity ERP is normal at onset (first hospitalization) of schizophrenia but becomes abnormal in the course of the disorder (this developing abnormality is associated with a loss of GM in auditory cortex). The mismatch negativity is thus potentially of use in tracking the ability of therapeutic interventions to minimize brain changes. It is not yet known if gamma abnormalities become evident early or late in the course of schizophrenia.
In recent years, the postpubescent period received increasing attention from researchers in the field of schizophrenia (Stevens, 2002). This interest stems largely from the fact that adolescence is associated with a significant rise in the risk for psychotic symptoms, particularly prodromal signs of schizophrenia (van Oel, Sitskoorn, Cremer, & Kahn, 2002; Walker, 2002). Further, rates of other psychiatric syndromes, including mood and anxiety disorders, escalate during adolescence. It has been suggested that hormonal changes may play an important role in this developmental phenomena, making ad olescence a critical period for the emergence of mental illness (Walker, 2002).
Puberty results from increased activation of the hypothalamic-pituitary-gonadal (HPG) axis, which results in a rise in secretion of sex hormones (steroids) by the gonads in response to gonadotropin secretion from the anterior pituitary. Rising sex steroid concentrations are associated with other changes, including increased growth hormone secretion.
There is also an augmentation of activity in the hypothalamic-pituitary-adrenal (HPA) axis during adolescence. This neural system governs the release of several hormones and is activated in response to stress. Cortisol is among the hormones secreted by the HPA axis, and researchers can measure it in body fluids to index the biological response to stress. Beginning around age 12, there is an age-related increase in baseline cortisol levels in normal children. The change from pre-to postpubertal status is linked with a marked rise in cortisol (Walker, Walder, & Reynolds, 2001) and a significant rise in cortisol clearance and in the volume of cortisol distribution.
The significance of postpubertal hormonal changes has been brought into clearer focus as researchers have elucidated the role of steroid hormones in neuronal activity and morphology (Dorn & Chrousos, 1997; Rupprecht & Holsboer, 1999). Neurons contain receptors for adrenal and gonadal hormones. When activated, these receptors modify cellular function and impact neurotransmitter function. Short-term effects (nongenomic effects) of steroid hormones on cellular function are believed to be mediated by membrane receptors. Longer-term effects (genomic effects) can result from the activation of intraneuronal or nuclear receptors. These receptors can influence gene expression. Brain changes that occur during normal adolescence may be regulated by hormonal effects on the expression of genes that govern brain maturation.
Gonadal and adrenal hormone levels are linked with behavior in adolescents. In general, both elevated and very low levels are associated with greater adjustment problems. For example, higher levels of the adrenal hormones (androstenedione) are associated with adjustment problems in both boys and girls (Nottelmann et al., 1987). Children with an earlier onset of puberty have significantly higher concentrations of adrenal androgens, estradiol, thyrotropin, and cortisol. They also manifest more psychological disorders (primarily anxiety disorders), self-reported depression, and parent-reported behavior problems (Dorn, Hitt, & Rotenstein, 1999). The relationship between testosterone and aggressive behavior is more pronounced in adolescents with more conflictual parent-child relationships, and this demonstrates the complex interactions between hormonal and environmental factors (Booth, Johnson, Granger, Crouter, & McHale 2003).
It is conceivable that hormones are partially exerting their effects on behavior by triggering the expression of genes that are linked with vulnerability for behavioral disorders. Consistent with this assumption, the heritability estimates for antisocial behavior (Jacobson, Prescott, & Kendler, 2002) and depression (Silberg et al., 1999) increase during adolescence. Further, the relationship between cortisol and behavior may be more pronounced in youth with genetic vulnerabilities. For example, increased cortisol is more strongly associated with behavior problems in boys and girls with fragile X than in their unaffected siblings (Hessl et al., 2002).
To date, there has been relatively little research on the HPG axis and schizophrenia, and there is no database on gonadal hormones in adolescent schizophrenia patients. The available reports on adult schizophrenia patients suggest that estrogen may serve to modulate the severity of psychotic symptoms and enhance prognosis (Huber et al., 2001; Seeman, 1997). Specifically, there is evidence that estrogen may have an ameliorative effect by reducing dopaminergic activity.
The role of the HPA axis in schizophrenia has received greater attention. A large body of research literature suggests a link between exposure to psychosocial stress and symptom relapse and exacerbation in schizophrenia (Walker & Diforio, 1997). It has been suggested that activation of the HPA axis mediates this effect (Walker & Diforio, 1997). Dysregulation of the HPA axis, including elevated baseline cortisol and cortisol response to pharmacological challenge, is often found in unmedicated schizophrenia patients (e.g., Lammers et al., 1995; Lee, Woo, & Meltzer, 2001; Muck-Seler, Pivac, Jakovljevic, & Brzovic, 1999). Patients with higher cortisol levels have more severe symptoms (Walder, Walker, & Lewine, 2000) and are more likely to commit suicide (Plocka-Lewandowska, Araszkiewicz, & Rybakowski, 2001).
Basic research has demonstrated that cortisol affects the activity of several neurotransmitter systems. This includes dopamine, a neurotransmitter that has been implicated in the etiology of schizophrenia (Walker & Diforio, 1997). The assumption is that increased dopamine activity plays a role in psychotic symptoms. Cortisol secretion augments dopamine activity. Thus it may be that when patients are exposed to stress and elevations in cortisol ensue, dopamine activity increases and symptoms are triggered or exacerbated.
Although there are no published reports on cortisol secretion in adolescents with schizophrenia, HPA axis function has been studied in adolescents with schizotypal personality disorder (Weinstein, Diforio, Schiffman, Walker, & Bonsall, 1999). Schizotypal personality disorder (SPD) involves subclinical manifestations of the symptoms of schizophrenia, including social withdrawal and unusual perceptions and ideas. This disorder is both genetically and developmentally linked with schizophrenia. The genetic link is indicated by the higher rate of SPD in the family members of patients diagnosed with schizophrenia. From a developmental perspective, there is extensive evidence that the defining symptoms of SPD often predate the diagnosis of schizophrenia, usually arising during adolescence.
When compared to healthy adolescents, adolescents with SPD show elevated baseline levels of cortisol (Weinstein et al., 1999) and a more pronounced developmental increase in cortisol when measured over a 2-year period (Walker et al., 2001). Further, SPD adolescents who show a greater developmental rise in cortisol are more likely to have an increase in symptom severity over time. This suggests that increased activation of the HPA axis may contribute to the worsening of symptoms as the child progresses through adolescence.
Research on the role of neurohormones in schizophrenia, especially the gonadal and adrenal hormones, should be given high priority in the future. In particular, it will be important to study hormonal processes in youth at risk for schizophrenia. There are several key questions to be addressed in clinical research. Are hormonal changes linked with the emergence of the prodromal phases of schizophrenia? Do rising levels of adrenal or gonadal hormones precede the onset of symptoms? Is there a relationship between hormonal factors and the brain changes that have been observed in the prodromal phase of schizophrenia? At the same time, basic science research is expected to yield new information about the impact of hormones on gene expression. This may lead to clinical research to explore the role of adolescent hormone changes on the gene expression in humans.
BRAIN CIRCUITRY IN SCHIZOPHRENIA
Information processing in the brain is a complex task, and even simple sensory information, such as recognizing a sight or a sound, engages circuits of cells in multiple regions of the brain. Scientists early in the 20th century imagined that brain function occurred in discrete steps along a linear stream of information flow. However, the recent emergence of brain imaging as an important tool for understanding the neuroscience of cognition and emotion has demonstrated that the brain operates more like a parallel processing computer with feed-forward and feedback circuitry that manages information in distributed and overlapping processing modules working in parallel. Thus, abnormal function in one brain region will have functional ripple effects in other regions, and abnormal sharing of information between regions, perhaps because of problems in the connectional wiring, can result in abnormal behavior even if individual modules are functionally intact.
In light of the elaborate and complex symptoms of schizophrenia, it is not surprising that researchers have increasingly focused on evidence of malfunction within distributed brain circuits rather than within a particular single brain region or module. Most of this work has been based on in vivo physiologic techniques, such as imaging and electrophysiology. At the same time, basic research in animals and to a lesser extent in humans has shown that the elaboration of brain circuitry is a lifelong process, especially the connection between cells in circuits within and between different regions of the cortex. This process of development and modification of connections between neurons is particularly dynamic during adolescence and early adult life. In this section, we will review some of the recent evidence that local and distributed abnormalities of brain circuitry are associated with schizophrenia and their implications for adolescent psychosis.
Two of the most often cited areas of the brain said to be abnormal in schizophrenia are the cortices of the frontal and temporal lobes. Indeed, damage to these regions caused by trauma, stroke, or neurological disease is more likely to be associated with psychosis than is damage to other brain regions. Recent studies using neuroimaging techniques have suggested that malfunction at the systems level—that is, at the relationship of processing in the temporal and frontal lobes combined—best characterizes the problem in patients with schizophrenia. For example, in a study of identical twins discordant for schizophrenia, differences within each twin pair in volume of the hippocampus predicted very strongly the difference in the function of the prefrontal cortex assayed physiologically during a cognitive task dependent on the function of the prefrontal cortex (Weinberger, Berman, Suddah, & Torrey, 1992).
A peculiar disturbance in the use of language, so-called thought disorder, is one of the cardinal signs of schizophrenia. Language is highly dependent on frontotemporal circuitry, which is disturbed in schizophrenia. When patients are asked to generate a list of words beginning with a specific consonant, instead of activating the frontal lobes and deactivating the temporal lobes, as seen in healthy subjects, they do the opposite. More detailed analyses have examined declarative memory encoding, storage, and retrieval as related to language. Encoding is manipulated by instructing subjects to process material more deeply, as, for example, to make semantic judgments about to-be-remembered words, such as whether the words represent living or nonliving, or abstract or concrete words. This deeper, more elaborate encoding is compared with a shallower, more superficial level of encoding, such as having subjects judge the font (upper case versus lower case) of each word presented. Compared with healthy controls, patients with schizophrenia show different patterns of fMRI activation for semantically encoded words, with significantly reduced left inferior frontal cortex activation but significantly increased left superior temporal cortex activation (Kubicki et al., 2003). During tests of word retrieval, patients with schizophrenia tend to show underengagement of the hippocampus, but at the same time their prefrontal cortex is overactive (Heckers et al., 1998). During performance of effortful tasks, by contrast, people with schizophrenia show increased activity in hippocampus and an alteration in the connection between hippocampus and anterior cingulate cortex (Holcomb et al., 2000; Medoff, Holcomb, Lahti, & Tamminga, 2001). These studies suggest that the information-processing strategy for encoding and retrieving learned information, which depends on an orchestrated duet between frontotemporal brain regions, is disturbed in patients with schizophrenia.
Similar results have been found in studies focused on prefrontal mediated memory, so-called working memory, in which the normal relationships between prefrontal activation and hip-pocampal deactivation are disrupted in schizophrenia (Callicott et al., 2000). Finally, recent statistical approaches to interpreting functional imaging results based on patterns of intercorrelated activity across the whole brain have demonstrated that abnormalities in schizophrenia are distributed across cortical regions. In particular, the pattern based on the normal rela-tionships between prefrontal and temporal cortical activity is especially abnormal (Meyer-Lindenberg et al., 2001). This apparent functional abnormality in intracortical connected-ness has been supported by anatomical evidence from diffusion tensor imaging, which has pointed to an abnormality in the WM links between frontal and temporal lobes (e.g., Kubicki et al., 2002).
The evidence for abnormal function across distributed cortical circuitry is quite compelling in schizophrenia, and other regions representing other circuits are also implicated (Tamminga et al., 2002; Weinberger et al., 2001). Indeed, it is not clear that any particular area of cortex is normal under all conditions. This may reflect simply the interconnectedness of the brain or it may suggest that schizophrenia is especially characterized by a “dysconnectivity.” It is impossible at the current level of our understanding of the disease to differentiate between these possibilities.
Schizophrenia disrupts not only circuitry linking brain regions but also the microcircuitry within brain regions, as shown by abnormal electrophysiologic activity during simple, early-stage “automatic processing” of stimuli, processing relatively independent of directed, conscious control. For example, healthy subjects automatically generate a robust EEG response in and near primary auditory cortex to tones differing slightly in pitch from others in a series (“mismatch” response), whereas the processing response in schizophrenia to the mismatch is much less pronounced (Wible et al., 2001).
Neurophysiological studies have focused largely on function of the cerebral cortex, but the pharmacological treatment of schizophrenia targets principally the dopamine system, which has long implicated the striatum and related subcortical sites. In fact, cortical function and activity of the subcortical dopamine system are intimately related, consistent with circuitry models of brain function. Animal studies have demonstrated conclusively that perturbations in cortical function, especially prefrontal function, disrupt a normal tonic brake on dopamine neurons in the brainstem, leading to a loss of the normal regulation of these neurons and to their excessive activation (Weinberger et al., 2001). It is thought that the prefrontal cortex helps guide the dopamine reward system toward the reinforcing of contextually appropriate stimuli. In the absence of such normal regulation, reward and motivation may be less appropriately targeted.
Neuroimaging studies of the dopamine system in patients with schizophrenia, particularly those who are actively psychotic, have found evidence of overactivity in the striatum (Laruelle, 2000). Recently, two studies reported that this apparent overactivation of the subcortical dopamine system is strongly predicted by measures of abnormal prefrontal cortical function (Bertolino et al., 2000; Meyer-Lindenberg et al., 2002). Moreover, reducing dopaminergic transmission with dopamine antagonists in subcortical dopamine-rich regions is associated with substantial alterations in frontal cortex function (Holcomb et al., 1996), presumably mediated through circuits connecting the striatum to the frontal cortex (Alexander & Crutcher, 1990). These data illustrate that what happens in the prefrontal cortex is very important to how other brain systems function and that the behavioral disturbances of schizophrenia involve dysfunction of diverse and interconnected brain systems.
Brain Circuitry and Implications for Adolescence
Contrary to long-held ideas that the brain was mostly grown-up after childhood, it is now clear that adolescence is a time of explosive growth and development of the brain. While the number of nerve cells does not change after birth, the richness and complexity of the connections between cells do, and the capacity for these networks to process increasingly complex information changes accordingly. Cortical regions that handle abstract information and that are critical for learning and memory of abstract concepts—rules, laws, codes of social conduct—seem to become much more likely to share information in a parallel processing fashion as adulthood approaches. This pattern of increased cortical information sharing is reflected in the patterns of connections between neurons in different regions of the cortex. Thus, the dendritic trees of neurons in the prefrontal cortex become much more complex during adolescence, which indicates that the information flow between neurons has become more complex (Lambe, Krimer, & Goldman-Rakic, 2000). The possibility that schizophrenia involves molecular and functional abnormalities of information flow in these circuits suggests that such abnormalities may converge on the dynamic process of brain maturation during adolescence and increase the risk of a psychotic episode in predisposed individuals.
Pathophysiology of Schizophrenia in Adolescence
Despite over a century of research, we have only a limited understanding of what causes schizophrenia and related psychotic disorders. Early studies of the biological basis of schizophrenia relied mostly on either postmortem studies of brains of people with this illness or brain imaging studies typically of older patients with chronic schizophrenia, many of whom were treated with medications. It was therefore difficult to know to what extent the observed changes were the results of aging, illness chronicity, or medication effects. One can avoid such difficulties by conducting studies of individuals in the early phases of schizophrenia (Keshavan & Schooler, 1992). First, these studies allow us to clarify which of the biological processes may be unique to the illness and which ones might be a result of medications or of persistent illness. Second, first-episode studies allow us to longitudinally evaluate the course of the brain changes, and how such changes can help us predict outcome with treatment. Follow-up studies suggest that less than half of early psychosis patients go on to develop a chronic form of schizophrenia with poor level of functioning and intellectual deficits (Harrison et al., 2001). An understanding of which patients may have such an outcome will greatly help treatment decisions early in the illness. Finally, not all who have features of the prodromal phases of the illness go on to develop the psychotic illness (Yung et al., 2003). Studies of the prodromal and early course of psychotic disorders provide an opportunity to elucidate the neurobiological processes responsible for the transition from the prodromal to psychotic phase of the illness.
Several conceptual models of the biology and causation of schizophrenia have been recently suggested, and serve to guide research into the early phase of this illness. One view, which dates back to the late 1980s, is the so-called early neurodevelopmental model (Murray & Lewis, 1987; Weinberger, 1987). This model posits abnormalities early during brain development (perhaps at or before birth) as mediating the failure of brain functions in adolescence and early adulthood. Several lines of evidence, such as an increased rate of birth complications, minor physical and neurological abnormalities, and subtle behavioral difficulties in children who later developed schizophrenia, support this view. However, many nonaffected persons in the population also have these problems; their presence cannot inform us with confidence whether or not schizophrenia will develop later in life. The fact that the symptoms typically begin in adolescence or early adulthood suggests that the illness may be related to some biological changes related to adolescence occurring around or prior to the onset of psychosis. Childhood is characterized by proliferation of synapses and dendrites, and normal adolescence is characterized by elimination or pruning of unnecessary synapses in the brain, a process that serves to make nerve cell transmission more efficient (Huttenlocher et al., 1982). This process could go wrong, and an excessive pruning before or around the onset phase of illness (Feinberg, 1982b; Keshavan, Anderson, & Pettegrew, 1994) has been thought to mediate the emergence of psychosis in adolescence or early adulthood. Our understanding of the underlying neurobiology of this phase of illness remains poor, however. Another view is that active biological changes could occur after the onset of illness, during the commonly lengthy period of untreated psychosis. This model proposes progressive neurodegenerative changes (Lieberman, Perkins, et al., 2001). It is possible that all three processes are involved in schizophrenia (Keshavan & Hogarty, 1999); additionally, environmental factors such as drug misuse (Addington & Addington, 1998) and psychosocial stress (Erickson, Beiser, Iacono, Fleming, & Lin, 1989) may trigger the onset and influence the course of schizophrenia. Careful studies of the early phase of schizophrenia can shed light on these apparently contrasting models. The three proposed pathophysiological models might reflect different critical periods for prevention and therapeutic intervention.
THE GENETICS OF SCHIZOPHRENIA
Remarkable progress has been made in understanding genetic factors related to schizophrenia. We will summarize this work in the following section. Since almost no work has been done specifically on the genetics of adolescent-onset schizophrenia, we focus on studies of typical samples of adult-onset cases.
Is Schizophrenia Familial?
The most basic question in the genetics of schizophrenia is whether the disorder aggregates (or “runs”) in families. Technically, familial aggregation means that a close relative of an individual with a disorder is at increased risk for that disorder, compared to a matched individual chosen at random from the general population. Twenty-six early family studies, conducted prior to 1980 and lacking modern diagnostic procedures and appropriate controls, consistently showed that first-degree relatives of schizophrenia patients had a risk for schizophrenia that was roughly 10 times greater than would be expected in the general population (Kendler, 2000). Since 1980, 11 major family studies of schizophrenia have been reported that used blind diagnoses, control groups, personal interviews, and operationalized diagnostic criteria. The level of agreement in results is impressive. Every study showed that the risk of schizophrenia was higher in first-degree relatives of schizophrenic patients than in matched controls. The mean risk for schizophrenia in these 11 studies was 0.5% in relatives of controls and 5.9% in the relatives of schizophrenics. Modern studies suggest that, on average, parents, siblings, and offspring of individuals with schizophrenia have a risk of illness about 12 times greater than that of the general population, a figure close to that found in the earlier studies.
Recently, results of the first methodologically rigorous family study of child-onset schizophrenia have been reported. Compared to parents of matched normal controls and children with ADHD, parents of childhood-onset schizophrenia had an over 10-fold increased risk for schizophrenia. This finding supports the hypothesis of etiologic continuity between childhood-and adult-onset schizophrenia (Asarnow, Tompson, & Goldstein, 2001).
To What Extent Is the Familial Aggregation of Schizophrenia Due to Genetic Versus Environmental Factors?
Resemblance among relatives can be due to either shared or family environment (nurture), to genes (nature), or to both. A major goal in psychiatric genetics is to determine the degree to which familial aggregation for a disorder such as schizophrenia results from environmental or genetic mechanisms. Although sophisticated analysis of family data can begin to make this discrimination, nearly all of our knowledge about this problem in schizophrenia comes from twin and adoption studies.
Twin studies are based on the assumption that “identical,” or monozygotic (MZ), and “fraternal,” or dizygotic (DZ), twins share a common environment to approximately the same degree. However, MZ twins are genetically identical, whereas DZ twins (like full siblings) share on average only half of their genes. Results are available from 13 major twin studies of schizophrenia published from 1928 to 1998 (Kendler, 2000). Although modest differences are seen across studies, overall, the agreement is impressive. Across all studies, the average concordance rate for schizophrenia in MZ twins is 55.8% and in DZ twins, 13.5%. When statistical models are applied to these data to estimate heritability (the proportion of variance in liability in the population that is due to genetic factors), the average across all 13 studies is 72%. This figure, which is higher than that found for most common biomedical disorders, means that, on average, genetic factors are considerably more important than environmental factors in affecting the risk for schizophrenia.
Adoption studies can clarify the role of genetic and environmental factors in the transmission of schizophrenia by studying two kinds of rare but informative relationships: (1) individuals who are genetically related but do not share their rearing environment, and (2) individuals who share their rearing environment but are not genetically related. Three studies conducted in Oregon, Denmark, and Finland all found significantly greater risk for schizophrenia or schizophrenia-spectrum disorders in the adopted-away offspring of schizophrenic parents than that for the adopted-away offspring of matched control mothers. The second major adoption strategy used for studying schizophrenia begins with ill adoptees rather than with ill parents and compares rates of schizophrenia between groups of biologic parents and groups of adoptive parents. In two studies from Denmark using this strategy, the only group with elevated rates of schizophrenia and schizophrenia-spectrum disorders were the biological relatives of the schizophrenic adoptees (Kety et al., 1994).
Twin and adoption studies provide strong and consistent evidence that genetic factors play a major role in the familial aggregation of schizophrenia. Although not reviewed here, evidence for a role for nongenetic familial factors is much less clear. Some studies suggest they may contribute modestly to risk for schizophrenia, but most studies find no evidence for significant nongenetic familial factors for schizophrenia.
What Psychiatric Disorders Are Transmitted Within Families of Individuals With Schizophrenia?
Since the earliest genetic studies of schizophrenia, a major focus of such work has been to clarify more precisely the nature of the psychiatric syndromes that occur in excess in relatives of schizophrenic patients. To summarize a large body of evidence, relatives of schizophrenia patients are at increased risk for not only schizophrenia but also schizophrenia-like personality disorders (best captured by the DSM-IV categories of schizotypal and paranoid personality disorder) and other psychotic disorders (Kendler, 2000). However, there is good evidence that relatives of schizophrenia patients are not at increased risk for other disorders, such as anxiety disorders and alcoholism. The most active debate in this area is the relationship between schizophrenia and mood disorders. Most evidence suggests little if any genetic relationship between these two major groups of disorders, but some research does suggest a relationship particularly between schizophrenia and major depression.
The evidence that other disorders in addition to schizophrenia occur at greater frequency in the close relatives of individuals with schizophrenia has led to the concept of the schizophrenia-spectrum—a group of disorders that all bear a genetic relationship with classic or core schizophrenia.
What Is the Current Status for Identifying Specific Genes That Predispose to Schizophrenia?
Given the evidence that genetic factors play an important role in the etiology of schizophrenia, a major focus of recent work has been to apply the increasingly powerful tools of human molecular genetics to localize and identify the specific genes that predispose to schizophrenia. Two strategies have been employed in this effort: linkage and association. The goal of linkage studies is to identify areas of the human genome that are shared more frequently than would be expected by relatives who are affected. If such areas can be reliably identified, then these regions may contain one or more specific genes that influence the liability to schizophrenia. The method of linkage analysis has been extremely successful in identifying the location of genes for simple, usually rare medical genetic disorders (termed “Mendelian” disorders) in which there is a one-to-one relationship between having the defective gene and having the disorder. This method, however, has had more mixed results when applied to disorders such as schizophrenia that are genetically “complex.” Such complex disorders are likely to be the result of multiple genes, none of which have a very large impact on risk, interacting with a range of environmental risk factors.
Eighteen genome scans for schizophrenia have been published between 1994 and 2002. None of these scans has revealed evidence for a single gene with a large impact on risk for schizophrenia. Indeed, these results suggest that the existence of a single susceptibility locus that accounts for a large majority of the genetic variance for schizophrenia can now be effectively ruled out.
The most pressing scientific issue in the interpretation of linkage studies of schizophrenia has been whether there is agreement at above-chance levels across studies on which individual regions of the genome contain susceptibility genes for schizophrenia. Until recently, the across-study agreement had not been very impressive.
Two recent findings have increased our confidence that linkage studies of schizophrenia may be producing reliable results. First, in a large-scale study of families containing two or more cases of schizophrenia, conducted in Ireland, the sample was divided, prior to analysis, into three random subsets (Straub, MacLean, et al., 2002). When a genome scan was performed on these three subsets, three of the four regions that most prominently displayed evidence for linkage (on chromosomes 5q, 6p, and 8p) were replicated across all three subsets. Interestingly, one region, on chromosome 10p, was not replicated even within the same study. Probably more important, Levinson and collaborators were able to obtain raw data from nearly all major published genome scans of schizophrenia to perform a meta-analysis—a statistical method for rigorously combining data across multiple samples (Lewis et al., 2003). Ten regions produced nominally significant results including 2q, 5q, 6p, 22q, and 8p. The authors concluded: “There is greater consistency of linkage results across studies than had been previously recognized. The results suggest that some or all of these regions contain loci that increase susceptibility to schizophrenia in diverse populations.”
On the Cusp of Gene Discovery in Schizophrenia
The evidence for replicated linkages in schizophrenia represents an important step toward the ultimate goal of identifying susceptibility genes and characterizing their biologic effects. Because the human genome contains within its 23 pairs of chromosomes over three billion nucleotides (i.e., “letters” in the genetic alphabet) and 30,000 genes (i.e., protein-encoding units), it is a large territory to explore. Linkage is a strategy to narrow the search and to provide a map of where the treasure (i.e., the genes) may lie. The linkage results in schizophrenia so far have highlighted several regions of the genome for a more thorough search. Association (also called linkage disequilibrium) is the next critical step in this search for the treasure. Linkage represents a relationship between regions of the genome shared by family members who also share the phenotype of interest—here, schizophrenia. It provides a low-resolution map because family members share relatively large regions of any chromosome. Association, however, represents a relationship between specific alleles (i.e., specific variation in a gene or in a genetic marker) and illness in unrelated individuals. It provides a high-resolution map because unrelated people share relatively little genetic information. For a given allele to be found more frequently in unrelated individuals with a similar disease than it is in the general population, the probability that this specific allele is a causative factor in the disease is enhanced. If the frequency of a specific allele (i.e., a specific genetic variation) is greater in a sample of unrelated individuals who have the diagnosis of schizophrenia than it is in a control population, the allele is said to be associated with schizophrenia. This association represents one of three possibilities: the allele is a causative mutation related to the etiology of the disease; the allele is a genetic variation that is physically close to the true causative mutation (i.e., in “linkage disequilibrium” with the true mutation); or the association is a spurious relationship reflecting population characteristics not related to the phenotype of interest. This latter possibility is often referred to as a population stratification artifact, meaning that differences in allele frequencies between the cases and control samples are not because of disease but because of systematic genetic differences between the comparison populations.
Association has become the strategy of choice for fine mapping of susceptibility loci and for preliminary testing of whether specific genes are susceptibility genes for schizophrenia. The strategy involves identifying variations (“polymorphisms”) in a gene of interest and then performing a laboratory analysis of the DNA samples to “type” each variation in each individual and determine its frequency in the study populations. Genetic sequence variations are common in the human genome and public databases have been established to catalog them. The most abundant sequence variations are single nucleotide polymorphisms (SNPs), which represent a substitution in one DNA base. Common SNPs occur at a frequency of approximately one in every 1,000 DNA bases in the genome and over two million SNPs have been identified. While SNPs are relatively common, most SNPs within genes either do not change the amino acid code or are in noncoding regions of genes (“introns”) and are thus not likely to have an impact on gene function.
Early association studies in schizophrenia focused on genes based on their known function and the possibility that variations in their function might relate to the pathogenesis of the disease. These so-called functional candidate gene studies had no a priori probability of genetic association. A number of studies compared frequencies of variations in genes related to popular neurochemical hypotheses about schizophrenia, such as the dopamine and glutamate hypotheses, in individuals with schizophrenia with those in control samples. In almost every instance the results were mixed, with some positive but mostly negative reports. Many of the positive studies were compromised by potential population stratification artifacts. However, because the effect on risk of any given variation in any candidate gene (e.g., a dopamine or glutamate receptor gene) is likely to be small (less than a twofold increase in risk), most studies have been underpowered to establish association or to rule it out.
Recent association studies have been much more promising, primarily because of the linkage results. Using the linkage map regions as a priori entry points into the human genetic sequence databases, genes have been identified in each of the major linkage regions that appear to represent at least some of the basis for the linkage results. Moreover, confirmation of association in independent samples have appeared, which combined with the linkage results comprise convergent evidence for the validity of these genetic associations. In the August and October 2002 issues of the American Journal of Human Genetics, the first two articles appeared that claimed to identify susceptibility genes for schizophrenia, starting with traditional linkage followed by fine association mapping. Both of these were in chromosomal regions previously identified by multiple linkage groups: dysbindin (DTNBP1) on chromosome 6p22.3 (Straub, Jiang, et al., 2002) and neuregulin 1 (NRG1) on chromosome 8p-p21 (Stefansson et al., 2002). Both groups identified the genes in these regions from public databases and then found variations (SNPs) within the genes that could be tested via an association analysis. In both studies, the statistical signals were strong and unlikely to occur by chance. In the January 2003 issue of the same journal, two further articles were published, replicating, in independent population data sets also from Europe, association to variations in the same genes (Schwab et al., 2003; Stefansson et al., 2003). In the December 2002 issue of the same journal, authors of a study on a large population sample from Israel reported very strong statistical association to SNPs in the gene for catechol-O-methyltransferase (COMT), which was mapped to the region of 22q that had been identified as a susceptibility locus in several linkage studies (Shifman et al., 2002). Positive association to variation in COMT had also been reported in earlier studies in samples from China, Japan, France, and the United States (Egan et al., 2001). Starting with the linkage region on chromosome 13q34, a group from France discovered a novel gene, called G72, and reported in two population samples association between variations in this gene and schizophrenia (Chumakov et al., 2002). The SNP variations in G72 have recently been reported to be associated with bipolar disorder as well.
In addition to these reports based on relatively strong linkage regions, several other promising associations have emerged from genes found in weaker linkage regions. For example, a weak linkage signal was found in several genome scans in 15q, a region containing the gene for the α–7 -nicotine receptor (CHRNA7; Raux et al., 2002). This gene has been associated with an intermediate phenotype related to schizophrenia, the abnormal P50 EEG evoked response. Preliminary evidence has been reported that variants in CHRNA7 are associated with schizophrenia as well. DISC-1 is a gene in 1q43, which was a positive linkage peak in a genome linkage scan from Finland. A chromosomal translocation originating in this gene has been found to be very strongly associated with psychosis in Scottish families having this translocation (Millar et al., 2000). Finally, in a study of gene expression profiling from schizophrenic brain tissue, a gene called RGS4 was found to have much lower expression in schizophrenic brains than in normal brains. This gene is found in another 1q region that was positive in a linkage scan from Canada, and SNPs identified in RGS4 have now been shown to be associated with schizophrenia in at least three population samples (Chowdari et al., 2002). This convergent evidence from linkage and association studies implicates at least seven specific genes as potentially contributing risk for schizophrenia.
From Genetic Association to Biological Mechanisms of Risk
Genetic association identifies genes but it does not identify disease mechanisms. Most of the genes implicated thus far are based on associations with variations that are not clearly functional, in the sense that they do not appear to change the integrity of the gene. Most are SNPs in intronic regions of genes, which do not have an impact on traditional aspects of gene function, such as the amino acid sequence or regulation of transcription. So, the associations put a flag on the gene but they do not indicate how inheritance of a variation in the gene affects the function of the gene or the function of the brain. More work is needed in searching for variations that may have obvious functional implications and in basic cell biology to understand how gene function affects cell function.
In two of the genes implicated to date, there is evidence of a potential mechanism of increased risk. Preliminary evidence suggests that SNPs in the promotor region of the CHRNA7 gene that are associated with schizophrenia affect factors that turn on transcription of the CHRNA7 gene, presumably accounting for lower abundance of CHRNA7 receptors, which has been reported in schizophrenic brain tissue (Leonard et al., 2002). This receptor is important in many aspects of hippocampal function and in regulation of the response of dopamine neurons to environmental rewards. Both hippocampal function and dopaminergic responsivity have been prominently implicated in the biology of schizophrenia. The COMT valine allele, which has been associated with schizophrenia in the COMT studies, translates into a more active enzyme, which appears to diminish dopamine in the prefrontal cortex. This leads to various aspects of poorer prefrontal function, in terms of cognition and physiology, which are prominent clinical aspects of schizophrenia, and to intermediate phenotypes associated with risk for schizophrenia (Weinberger et al., 2001). The COMT valine allele also is associated with abnormal control of dopamine activity in the parts of the brain where it appears to be overactive in schizophrenia (Akil et al., 2003). Thus, inheritance of the COMT valine allele appears to increase risk for schizophrenia because it biases toward biological effects implicated in both the negative and positive symptoms of the illness.
Schizophrenia-Susceptibility Genes and Adolescence
It is not obvious how the genes described would specifically relate to adolescence and the emergence of schizophrenia during this time of life. The evidence so far suggests that each of the candidate susceptibility genes has an impact on fundamental aspects of how a brain grows and how it adapts to experience. Each gene may affect the excitability of glutamate neurons—directly or through GABA neuron intermediates, and indirectly through the regulation of dopamine neurons by the cortex. These are fundamental pro cesses related to the biology of schizophrenia. These are also processes that may be especially crucial to adolescence because cortical development and plasticity are changing dramatically during this period. Thus, it is conceivable that the variations in the functions of these genes associated with schizophrenia lead to compromises and bottlenecks in these processes.
The Potential Gene-Finding Utility of Intermediate or Endophenotypes
Despite encouraging results from recent linkage and association studies, the literature also contains prominent failures and inconsistencies. Failures to replicate linkage and association signals for schizophrenia suggest that genomic strategies may benefit from a redirection based on our current understanding of the pathophysiology of schizophrenia. For example, the power of genetic studies may increase by examining linkage with quantitative traits that relate to schizophrenia rather than with a formal diagnosis itself. The concept of using intermediate phenotypes, or endophenotypes, is not new (Gottesman & Gould, 2003), but has only recently started to enjoy widespread popularity among those seeking genes for schizophrenia. Gottesman and Shields suggested over 30 years ago that features such as subclinical personality traits, measures of attention and information processing, or the number of dopamine receptors in specific brain regions might lie “intermediate to the phenotype and genotype of schizophrenia” (Gottesman & Shields, 1973). Today, other traits, such as eye-movement dysfunctions, altered brain-wave patterns, and neuropsychological and neuroimaging abnormalities, are under consideration as potentially useful endophenotypes of schizophrenia, because all of these are more common or more severe in schizophrenic patients and their family members than in the general population or among control subjects (Faraone et al., 1995). These deficits may relate more directly than the diagnosis of schizophrenia to the aberrant genes. At the biological level, this is a logical assumption, as genes do not encode for hallucinations or delusions; they encode primarily for proteins that have an impact on molecular processes within and between cells. Thus, endophenotypes may serve as proxies for schizophrenia that are closer to the biology of the underlying risk genes.
Early Findings from Molecular Genetic Studies of Endophenotypes
While much recent work has been dedicated toward establishing the heritability of endophenotypes, only a handful of molecular genetic studies of endophenotypes have emerged. Results observed to date have been encouraging, in that some chromosomal loci that have been found to harbor genes for schizophrenia have also shown evidence for linkage with an endophenotype. For example, linkage with an auditory-evoked brain wave pattern (the P50 endophenotype) has been observed independently in two samples of schizophrenia pedigrees on chromosome 15 at the locus of the α–7 -nicotinic receptor gene, where some evidence for linkage had previously been observed using traditional diagnostic classifications (Leonard et al., 1996; Raux et al., 2002). However, the greater potential of endophenotype studies is that genes might be identified that would not be implicated from regions of the genome highlighted in linkage regions. This is because minor genes for schizophrenia may turn out to be major genes for some index of central nervous system dysfunction. The proof of this has been supported by evidence that COMT, which is a weak susceptibility gene for schizophrenia, is a relatively strong factor in normal human frontal lobe function (Weinberger et al., 2001).
Whether classical criteria or quantitative phenotypes are used to further study schizophrenia, refining the definition of an “affected” individual is a top priority for genetic studies. Because not all individuals with schizophrenia-susceptibility genes develop the actual disorder, understanding the measurable effects of these aberrant genes is a critical step in tracking their passage through affected pedigrees and in identifying their clinical biology. In the near future, the amount and types of expressed protein products of these disease genes may be used as the ultimate endophenotype for schizophrenia. To the extent that we can reduce measurement er ror and create measures that are more closely tied to individual schizophrenia genes, we will greatly improve our understanding of the genetics of schizophrenia.
Genetic Counseling Issues and Schizophrenia
With increasing attention in the media to issues relating to genetics and particularly the role of genetic factors in mental illness, an increasing number of individuals will likely be seeking genetic counseling for issues related to schizophrenia. In our experience, by far the most common situation is a married couple who are contemplating having children and the husband or wife has a family history of schizophrenia. They typically ask any combination of three questions: First, is there a genetic test that can be performed on us to determine whether we have the gene for schizophrenia and whether we might pass it on to our children? Second, is there an in utero test that can be given that would determine the risk of the fetus to develop schizophrenia later in life? Third, what is the risk for schizophrenia to our children?
Unfortunately, given the current state of our knowledge, answers to the first two questions are no, we are not yet in the position of having a genetic test that can usefully predict risk for schizophrenia. We would also often add a statement to the effect that this is a very active area of research and there is hope that in the next few years, some breakthrough might occur that would allow us to develop such a test. But, right now we really do not know when or even if that will be possible.
By contrast, useful information can be provided for the third question. Most typically, the husband or wife has a parent or sibling with schizophrenia and they themselves have been mentally healthy. Therefore, the empirical question is what is known about the risk of schizophrenia to the grandchild or niece or nephew of an individual with schizophrenia. Interestingly, this is a subject that has not been systematically studied since the early days of psychiatric genetics in the first decades of the 20th century. The results of these early studies have been summarized in several places, most notably by Gottesman (Gottesman & Shields, 1982), with aggregate risk estimates for schizophrenia of 3.7% and 3.0%, respectively, in grandchildren or nieces and nephews of an individual with schizophrenia. However, this is a considerable overestimate if the parent with the positive family history remains unaffected. That is, the risk to a grandchild or niece or nephew of an individual with schizophrenia when the intervening parent never develops the illness is probably under 2%. Most individuals find this information helpful and broadly reassuring.
By the time this chapter is read, a great deal more information is likely to have accumulated about the scientific status of these findings. At this early stage, several trends are noteworthy. First, including unpublished reports known to the authors, at least some of these potential gene discoveries have now been replicated enough times that it is increasingly unlikely that they are false-positive findings (due, for example, to the performance of many statistical tests). Second, we can expect that the biochemical pathways represented by these genes will be explored at the level of basic cell biology and new leads about pathogenesis and potential new targets for prevention and treatment will be found. Third, we can expect a number of studies to emerge that will try to understand whether expression of these genes are changed in the brains of schizophrenia patients. Fourth, efforts are already under way to try to understand how these genes influence psychological functions such as attention, sensory gating, and memory that are disturbed in schizophrenia. Fifth, intense efforts will be made to try to determine whether these different genes are acting through a common pathway as, for example, has been postulated for the four known genes for Alzheimer's disease.