|
Book Title: Treating and Preventing Adolescent Mental Health Disorders
> pp. [140]-[144]
UNDEFINED: AUTHORS
Treating and Preventing Adolescent Mental Health Disorders
Print ISBN 9780195173642, 2005
pp. [140]-[144]
sensitive than others, and not having a steady girlfriend. Cannon et al. ( 1997) also noted the same relationship. Another twist to the story about premorbid behavioral differences comes from the recent recognition that some of the individual parts of the schizophrenia syndrome, such as hallucinations or delusions, can exist in otherwise well-functioning individuals in the population. However, they are indeed associated with greater risk of occurrence of subsequent schizophrenia whether they occur in early adolescence (Poulton et al., 2000) or adulthood (Myin-Germeys, Krabbendam, Delespaul, & Van Os, 2003). Thus, there seems to be a consistency over childhood and adolescence and across several types of study regarding the presence of premorbid behavioral differences. People who will develop schizophrenia as adolescents and adults are different from their peers in terms of behavior in childhood, just as Bleuler noted a century ago; the effects may be even more widespread than he thought.
Cognitive Function and IQ
This aspect of psychological function also shows differences in the premorbid period. Aylward,
Walker, and Bettes ( 1984) have provided a comprehensive review of intelligence in schizophrenia. They concluded that intellectual function is lower in prepsychotic individuals than in age-matched controls. Linking the prepsychotic deficit to outcome, they raised the question as to whether IQ may be an independent factor that can protect otherwise vulnerable individuals, or whether the deficits are part of that vulnerability. Once again, the birth cohort studies shed light on the question. Cannon et al. ( 2002) showed that mean IQ test scores were consistently lower during childhood in those children who developed schizophreniform disorder (Fig. 7.5). This mean shift in premorbid IQ was also seen in two British cohorts (Jones & Done, 1997). When the childhood IQ data from the 1946 cohort (Pidgeon, 1964, 1968) is studied in greater detail, it is clear that the lower mean premorbid IQ is not due to a subset of people with very low scores; rather, the whole distribution of those who develop schizophrenia when they reach adolescence or adulthood is shifted down—most children seem not to be doing as well as they might have been expected to perform (Jones, Rodgers, Murray, & Marmot, 1994). This is a similar situation to the motor findings in the Finnish cohort (Cannon et al., 1999). It is not that there is a
end p.140
PRINTED FROM Treating and Preventing Adolescent Mental Health Disorders (www.oup.com/amhi-treatingpreventing)
© Copyright Oxford University Press, 2006. All Rights Reserved
group of very abnormal individuals driving the findings; the effects are seen across the normal range. David, Malmberg, Brandt, Allebeck, and Lewis ( 1997; see above) replicated this result in the Swedish conscript study, although the measures were later in life at age 18. There was no evidence of a threshold effect below or above which this relationship did not hold. Very bright individuals can develop schizophrenia, but they are less likely to than those who are less able. Put another way, any individual is more likely to develop schizophrenia than someone who is more able in terms of IQ, although the effect is small. Recent interest in the cognitive aspects of schizophrenia (David & Cutting, 1994; Green, 1998) suggests a parsimonious conclusion that prepsychotic IQ deficits (and perhaps social characteristics) may be manifestations of the same abnormal cognitive processes that later result in psychosis.
What Do Premorbid Abnormalities Mean?
The range of differences in the developmental histories of people who develop schizophrenia when they are older suggests that something to do with the causes of this syndrome is active long before the characteristic features begin (Marenco & Weinberger, 2000). There is evidence for many such early factors, including genetic effects (Jones & Murray, 1991; Fish et al., 1992), obstetric complications (Cannon et al., 2002), psychosocial stresses, famine, infections, and other toxic events during brain development (see Jones, 1999, for review). It seems that many events that may lead to early brain development being suboptimal may increase the risk of later schizophrenia. There may be specific causes or combinations of causes, such as gene–environment interactions, that make people vulnerable to developing the schizophrenia syndrome, perhaps after later, necessary events that act as triggers. These may include normal (Weinberger, 1987, 1995) or abnormal brain development (Feinberg, 1982a, 1997; Pogue-Geile, 1997), as well as traditional precipitants such as psychosocial stressors or drugs (see Chapter 6). The behavioral, motor, language, and cognitive differences shown in the premorbid period may be manifestations of vulnerability or predisposition to schizophrenia; they may not be risk modifiers in themselves. These indicators seem remarkably homogeneous—in retrospect, like a final common pathway. The idea of only a subgroup of individuals having this manifest vulnerability, as suggested in the seminal views of developmental aspects of schizophrenia (Murray & Lewis, 1987), is not supported by recent research. Most people or even every person who develops the syndrome may have had a degree of developmental vulnerability, although this will not have been obvious at the time. The early motor findings in the Finnish birth cohort (Fig. 7.4) are consistent with the vulnerability being due to developmental processes being generally less efficient, the formation or enhancement of functional neural networks, for instance. The greater the inefficiency, the greater the risk of schizophrenia when that same inefficiency is played out in the formation of complex and integrative systems later in adolescence and adult life (Chapter 5). There are several candidates to explain this unifying vulnerability. These include hormonal events (Walker & Bollini, 2002) that are able to tie together motor and other system abnormalities in early life and links with psychosocial stress in models of predisposition and precipitation (Walker, Lewis, Loewy, & Palyo, 1999). Molecular biology and the investigation of not only the presence but also the functional activity of genes and the proteins that code them may yield other dimensions of vulnerability. For instance, Tkachev et al. ( 2003) showed that expression of genes associated with glial cells facilitate the nutritional support of nerve cells (oligodendrocytes) and with myelin, the insulating sheaths provided for these neurons, were down-regulated in the frontal cortex of brains of deceased people who had suffered from schizophrenia. Expression seems a very good candidate for the homogeneous vulnerability factor posited in this account of premorbid abnormalities before schizophrenia, and may be an endophenotype or hidden manifestation of the disorder. The deficient gene expression remains to be
end p.141
PRINTED FROM Treating and Preventing Adolescent Mental Health Disorders (www.oup.com/amhi-treatingpreventing)
© Copyright Oxford University Press, 2006. All Rights Reserved
demonstrated before onset of schizophrenia and will itself have its own prior causes. As mentioned at the beginning of this section, premorbid features of schizophrenia are not yet of use in terms of prediction and early intervention. They occur in multiple domains, but many of the effects we can measure are subtle and leave individuals remaining well within the wide range of normality. Premorbid features tell us a great deal about what we should be looking for in terms of underlying mechanisms and causes of schizophrenia and when these may operate; these features are signposts toward these mechanisms. As we learn about the processes that underpin the behavioral, cognitive, and motor differences that we can measure in the premorbid phase of schizophrenia, we may become able to identify those individuals who are vulnerable with enough precision to be able to do something useful for them.
Developmental Precursors of Adolescent-Onset Schizophrenia
There are precursors of schizophrenia prior to the first onset of psychosis in many but not all adolescents who develop schizophrenia. As will be seen below, the precursors of schizophrenia can be subtle changes in basic brain functions such as motor functions, attention and memory, certain behavior problems, or attenuated schizophrenic symptoms. Identifying the developmental precursors of adolescent-onset schizophrenia has important implications for both enhancing our understanding of the underlying neurobiology of schizophrenia and developing of preventive interventions for schizophrenia. Neurobiological factors present in individuals at high risk for developing a schizophrenic disorder, prior to the onset of frank psychotic symptoms, may represent potential etiological factors for schizophrenia. A number of brain systems known to be disturbed in schizophrenia, including prefrontal and medial temporal lobes (Selemon & Goldman-Rakic, 1999; Weinberger, 1986), may underlie certain neurocognitive impairments in children at risk for schizophrenia (Asarnow, 1983; Cannon et al., 1993). Determining how these neurobiological factors evolve
when a schizophrenic disorder develops could provide important clues about how the diathesis for schizophrenia is potentiated into the overt disorder. A combination of disease-related progressions and maturational changes are hypothesized to exacerbate these dysfunctions when individuals at risk for the disorder convert to having the disorder.
Research Methods Used to Identify Developmental Precursors
Two broad classes of methods have been used to identify developmental precursors of schizophrenia. The first class of methods is prospective studies of children. A common feature of prospective methods is identifying, then characterizing, a group of children and following them up to determine which children subsequently develop a schizophrenic disorder. One important prospective method is to study children who are at increased statistical risk of developing a schizophrenic disorder. The lifetime risk for schizophrenia in the general population is less than 1%. Very large samples are required to prospectively identify the precursors of schizophrenia by following up children drawn from the general population. Given the population base rate of schizophrenia (<1%), one would need to start off with at least 2,500 children (without accounting for subjects being lost to follow-up) to identify the developmental precursors of schizophrenia in 25 individuals. High-risk studies ascertain individuals with an increased lifetime risk for schizophrenia for inclusion in prospective, longitudinal studies. This is typically accomplished by studying the children of parents with schizophrenia. The lifetime risk for schizophrenia for children of one parent with schizophrenia is approximately 10% to 12%, an approximately 10-fold increase in risk for the disorder. High-risk studies frequently measure putative etiological factors for schizophrenia prior to the onset of the disorder. In this way, studies of children at risk for schizophrenia provide a vehicle for testing hypotheses about etiological factors in schizophrenia. Most (85% to 90%) patients with schizophrenia do not have parents with a schizophrenic disorder. This has raised the concern that findings
end p.142
PRINTED FROM Treating and Preventing Adolescent Mental Health Disorders (www.oup.com/amhi-treatingpreventing)
© Copyright Oxford University Press, 2006. All Rights Reserved
from “genetic high-risk” samples may not accurately describe the developmental precursors of schizophrenia in the much larger number of individuals who develop schizophrenia but do not have a schizophrenic parent. Recognition of this problem has led to an interest in complementary strategies for identifying developmental precursors of schizophrenia. Birth cohort studies are prospective studies that can provide information on precursors of schizophrenia but do not have some of the ascertainment biases inherent in high-risk studies. In contrast to studies of children at risk for schizophrenia, birth cohort studies follow up large, representative samples of entire birth cohorts. Birth cohort studies are designed to provide information about a wide range of medical, psychiatric, and social conditions, so they use very large samples, literally thousands of subjects. For example, the 1946 British birth cohort study that provided im-portant data on developmental precursors of schizophrenia studied almost 5,400 children born during the week of March 9, 1946, then systematically followed them up to determine that 30 children developed schizophrenia as well as a broad range of other psychiatric and medical outcomes. A great strength of birth cohort studies is the large, representative sample size. However, a limitation of birth cohort studies is that because they are not typically designed to test hypotheses about any particular disorder, they use a rather broad range of measures, which are not specifically tailored to measure potential precursors of schizophrenia. By studying children prior to the onset of the disorder it becomes possible to identify the precursors or antecedents of the disorder, not the consequences of the disorder—for example, the initiation of antipsychotic drug treatment. We will review some of the key findings that have emerged from three decades of studies of children at risk for schizophrenia and from birth cohort studies. A second class of methods involves the collection of information on the premorbid development of individuals, usually adults, who have been diagnosed with schizophrenia. Some of the earliest studies of this type relied on retrospective reports from informants who knew the patient as a child. This approach has obvious limita
tions, among them being that recollections of the past may be subject to bias. The follow-back method features the ascertainment of individuals with schizophrenia and then, using different types of archival material, characterization of them prior to the onset of psychosis. Since the focus of this section is on adolescent-onset schizophrenia, we will emphasize the few studies that ascertained adolescent-onset schizophrenics. Follow-back studies vary in the type of archival material used to describe the premorbid characteristics of individuals who develop schizophrenia. There is wide agreement (see Watt, Grubb, & Erlenmeyer-Kimling, 1982) about the advantages of using contemporaneous childhood records over retrospective interviews to reconstruct the premorbid histories of individuals who develop schizophrenia. The major limitation of follow-back studies is that the childhood evaluations were not guided by specific hypotheses about the age-specific manifestations of schizophrenia, and as a consequence, the most informative measures may not have been collected. These studies also have ascertainment biases, the nature of which varies depending on how the sample of schizophrenia patients was identified. Birth cohort and follow-back studies can show associations between childhood characteristics and the development of schizophrenia because in both types of studies, individuals with schizophrenia have been identified. These associations are prospective in birth cohort studies, and retrospective in follow-back studies. Because the data used to describe childhood risk factors in birth cohort and follow-back studies were not collected with the intent of testing hypotheses about schizophrenia, the measures may not be sensitive to some of the more subtle manifestations of liability to schizophrenia. In contrast, the measures included in more recent studies of children at risk for schizophrenia were specifically designed to tap vulnerability to schizophrenia. Most studies of children at risk for schizophrenia, while intended to be longitudinal, were not able to follow up subjects through the age of risk to determine which high-risk subjects developed a schizophrenic disorder. Consequently, although there are extensive cross-sectional
end p.143
PRINTED FROM Treating and Preventing Adolescent Mental Health Disorders (www.oup.com/amhi-treatingpreventing)
© Copyright Oxford University Press, 2006. All Rights Reserved
comparisons of children at risk for schizophrenia to controls, there are far fewer data on the long-term predictive validity of childhood risk factors identified in high-risk studies. If the results of follow-back studies of adolescent-onset schizophrenia patients yield converging results to those of children at risk for schizophrenia and birth cohort studies, then the generalizability and validity of the results will hold more value for future research and treatment.
A Developmental Perspective on Risk Factors
There are relatively age-specific manifestations of liability to schizophrenia (see J. Asarnow, 1988; R. Asarnow, 1983; Erlenmeyer-Kimling et al., 2000; Walker, 1991 for reviews), and the manifestations to liability to schizophrenia are somewhat different at different ages. For example, one of the interesting findings that emerges from a review of developmental precursors of schizophrenia is that some deficits observed during infancy that are frequently found in high-risk, birth cohort, and follow-back studies are not found in later stages of development. Another important reason to attend to the developmental progression of risk factors is that, from the point of view of targeting individuals for prevention, risk factors more proximal to the period of time when schizophrenia develops may have better diagnostic accuracy than, for example, infancy predictors. Table 7.1 summarizes some of the major findings concerning precursors of schizophrenia at three different developmental periods: infancy, early childhood, and middle childhood and early adolescence. Table 7.1 is not an exhaustive summary of the results of high-risk, birth cohort, and follow-back studies. Rather, Table 7.1 presents the characteristics that best differentiate high-risk children from controls or predict later development of schizophrenia that have thus far been identified in the literature. Cited below are comprehensive reviews of the results of high-risk, birth cohort, and follow-back studies. The format of Table 7.1 was modeled after a review by J. Asarnow ( 1988), and the entries for
studies on high-risk children come from reviews by J. Asarnow ( 1988), Erlenmeyer-Kimling ( 2000, 2001), R. Asarnow ( 1983), and Cornblatt and Obuchowski ( 1997). The entries for birth cohort studies are based on reviews by Jones, Rogers, Murray, and Marmot ( 1994) and by Jones and Tarrant ( 1999). The data for entries of follow-back studies of adolescent-onset schizophrenia come from Watkins, Asarnow, and Tanguay ( 1988) and Walker, Savoie, and Davis ( 1994). Watt and Saiz ( 1991) provided a broad review of follow-back studies of adult-onset schizophrenia. Two types of risk characteristics are differentiated into separate columns in Table 7.1: endophenotypes versus clinical and behavioral features. Endophenotypes are putative reflections of the underlying schizophrenic genetic diathesis. Most of the putative endophenotypes employed in high-risk studies are neuromotor or neurocognitive functions (e.g., language, attention, and memory) believed to tap central nervous system disturbances that reflect liability to schizophrenia. In contrast, clinical and behavioral features are either nonschizophrenic psychiatric symptoms or behavior problems which, while they may reflect the underlying genetic diathesis, are much more proximal to the overt symptoms of schizophrenia. The reason for making this distinction is that these two different classes of risk characteristics have somewhat different implications as targets for prevention.
The results of high-risk studies have to be considered in the context of a major limitation: there are limited data on how well the cross-sectional differences between children at risk for schizophrenia and matched controls predict the later onset of schizophrenia. Only six studies of children at risk for schizophrenia have obtained diagnostic evaluations in adulthood or late adolescence: (1) the New York High-Risk study (Fish, 1984); (2) the Copenhagen High-Risk project (Cannon et al., 1993; Mednick & Schulsinger, 1968); (3) the Israeli High-Risk study (Ingraham, Kugelmass, Frankel, Nathan, & Mirsky, 1995); (4) the New York High-Risk project (Erlenmeyer-
end p.144
doi:10.1093/9780195173642.003.0008
|
|
|
|
![Figure 7.5 Mean standardized scores for IQ performance at four ages during childhood for adults who indicated symptoms of schizophreniform disorder (36), mania (20), or anxiety/depression (278), compared with controls (642) [from Cannon, M., Caspi, A., Moffitt, T.E., Harrington, H., Taylor A., & Murray, R.M. (2002). Evidence for early-childhood, pan-developmental impairment specific to schizophreniform disorder: results from a longitudinal birth cohort. Archives of General Psychiatry, 59, 449–456. Copyright © 2002, American Medical Association. All rights reserved, used with permission].](/anbrg/public/mentalhealth/9780195173642/figures/graphic013.gif)
