|
1
| 2 | 3
| 4 | 5
Psychosocial Factors
Kinney and others found elevations on the Thought Disorder Index
(TDI) in biological relatives of schizophrenia patients, compared
with normal control subjects (35). In contrast, elevations on the
TDI were not evident in the adoptive relatives of either schizophrenia
patients or control subjects. Consistent findings have been reported
by others (36–39). Wahlberg and others showed that young adult
offspring of mothers with schizophrenia were more likely to be thought-disordered
when raised by adoptive mothers who showed “communication deviance,”
whereas adoptees raised by adoptive parents with low communication
deviance were less likely to show thought disorder (38). In contrast,
control subject adoptees born from mothers without schizophrenia
showed no relation between adoptee thought disorder and communication
deviance in the adoptive parents. Thus, communication deviance appears
to be a risk factor only for people with a genetic vulnerability
to schizophrenia.
Pregnancy and Delivery Complications
Many studies have found an elevated prevalence of pregnancy and
delivery complications surrounding the births of individuals who
eventually developed schizophrenia. For example, schizophrenia patients
are more likely to have been born prematurely and to have had relatively
low birth weights (40,41). Other studies have confirmed a relation
between various pregnancy and delivery complications and subsequent
schizophrenia (42,43), but McNeil (44) suggested, and others confirmed
(45–48), that complications leading to oxygen deprivation or
trauma might be the most relevant to the subsequent development
of schizophrenia.
Zornberg and others studied 693 adults born between 1959 and 1966
and followed up for an average of 19 years (48). Their experience
of fetal and neonatal complications had been recorded at the time
of their birth. Hypoxic ischemia-related complications predicted
a doubling of the risk for psychosis, especially for psychoses not
associated with mood disorders. Nonpsychotic mood disorders were
unrelated to hypoxic ischemia–related complications. The data
show a strikingly elevated risk for schizophrenia and other nonaffective
psychoses associated with this classification of antecedent hypoxic
ischemia–related fetal or neonatal complications.
Several studies have found that pregnancy and delivery complications
tend to be more common in sporadic schizophrenia patients (that
is, those having no affected close relatives) than in familial schizophrenia
patients (49–52). McNeil and others found that pregnancy and
delivery complications among MZ twin pairs were more common among
those discordant for schizophrenia than among those concordant for
schizophrenia (53). A similar trend has been reported by Onstad
and others (54). Further, among MZ twins discordant for schizophrenia,
pregnancy and delivery complications were more likely to involve
the twin with schizophrenia than the cotwin without schizophrenia
(55). However, a review of 6 systematically ascertained twin samples
found no differences in birth weight between schizophrenic and well
MZ cotwins (56).
It is also possible that pregnancy and delivery complications merely
activate the genetic predisposition to schizophrenia. For example,
DeLisi and others found more pregnancy and delivery complications
among siblings with schizophrenia than among siblings without schizophrenia
in families having at least 2 siblings with the disorder (57). This
suggests that pregnancy and delivery complications work in combination
with familial factors. Similarly, among children of mothers with
schizophrenia, pregnancy and delivery complications predict subsequent
psychiatric abnormality (58). Among these high-risk children, the
children who developed schizophrenia had been exposed to more pregnancy
and delivery complications (59) This work also identified a group
of high-risk children who had the least complicated births. Because
these children had milder schizophrenia-like conditions but not
psychosis, the investigators suggested that schizophrenia might
result from the interaction of susceptibility genes with environmental
events such as pregnancy and delivery complications. In this model
of etiology, neither risk factor alone is sufficient for the development
of the disorder.
Viruses
The idea that schizophrenia is caused by a virus is consistent
with epidemiologic and clinical observations (60). Schizophrenia
patients are more likely to have been born during the late winter
and spring months, when the fetus is at increased risk for exposure
to viruses (61–63). Several indices of heightened immune-system
responsiveness are apparent in schizophrenia patients, including
elevations in herpes antibody titre, immunoglobulins, cytomegalovirus
antibody titre, interleukin-2 receptors, alpha interferon, and autoantibodies
(64–70). Of note, an allelic association between schizophrenia
and the human leukocyte antigen (HLA) A9 locus has been reported
(71). Since the HLA loci are known to be associated with autoimmune
diseases, these data led Wright and others to suggest that some
cases of schizophrenia may have an autoimmune basis (72).
Mednick and others studied a Finnish cohort who were exposed to
the influenza epidemic of 1957 at various stages of their fetal
life (73). Those exposed to the epidemic during their second trimester
of development were at increased risk for subsequent schizophrenia.
However, a Scottish study failed to find an increased risk for schizophrenia
associated with influenza epidemics in 1918, 1919, and 1957. Analyses
limited to the city of Edinburgh in 1957 supported the viral hypothesis,
but the nationwide data did not (74). Further, only limited evidence
of an association between viral epidemics and schizophrenia was
found in an American study (75).
Barr and others criticized the negative studies on methodological
grounds and replicated the Finnish results in a Danish sample (76).
Takei and others found that female subjects but not male subjects
exposed to influenza epidemics 5 months prior to birth had an increased
rate of schizophrenia in adulthood (77). Sham and others found a
link between schizophrenia and maternal influenza but suggested
that it would account for less than 2% of schizophrenia cases (78).
In contrast, Crow and Done did not find evidence to support the
hypothesis that maternal influenza was a risk factor for schizophrenia
(79).
Several studies report that nonfamilial schizophrenia patients
are more likely to have been born during the winter months (and
hence, are more likely to have been exposed to viruses), compared
with familial schizophrenia patients (80–83). For example,
Roy and others reported that 32% of sporadic schizophrenia patients
were born between December 21 and March 21, compared with 18% of
familial patients (82). Goldstein and others delineated a cluster
of schizophrenia patients characterized by poor premorbid history,
flat affect, winter birth, and no family history of schizophrenia
(80). This cluster was enriched with male sporadic schizophrenia
cases.
Other studies have implicated winter or spring births as a risk
factor for familial schizophrenia. For example, Pulver and others
found higher rates of schizophrenia among relatives of probands
born between February and May. The lowest rates among relatives
occurred for probands born between October and January. This effect
was observed for all relatives of female probands but only for younger
relatives of male probands. The older relatives of male probands
were actually less likely to have schizophrenia if the proband had
been born between February and May (84).
Buka and others studied the association between maternal cytokines
in pregnancy and the subsequent development of psychotic disorders
in offspring (85). Their sample included mothers of 27 adults with
psychotic disorders and 50 matched control subjects. Using enzyme
immunoassay, they analyzed serum samples for interleukin-1 beta
(IL-1beta), interleukin-2 (IL- 2), interleukin-6 (IL-6), interleukin-8
(IL-8), and tumour necrosis factor alpha (TNF-alpha). Maternal levels
of TNF-alpha were significantly elevated among the adults with psychotic
disorders. Higher maternal cytokine levels predicted an increased
likelihood of psychosis. They did not find significant differences
between case and control mothers in the serum levels of IL- 1beta,
IL-2, IL-6, and IL-8.
In another study of the same sample, Buka and others reported that
maternal levels of IgG- and IgM-class immunoglobulins taken prior
to the delivery of the child were significantly elevated among the
offspring with psychosis (86). They also found a significant association
between maternal antibodies to herpes simplex virus type 2 glycoprotein
gG2 and subsequent psychosis in offspring. They did not find significant
differences between case and control mothers either in the serum
levels of IgA-class immunoglobulins, or in specific IgG antibodies
to herpes simplex virus type 1, cytomegalovirus, toxoplasma gondii,
rubella virus, human parvovirus B19, Chlamydia trachomatis, or human
papillomavirus type 16.
A Typology of Prevention Approaches Based on Target Population
The terms “universal,” “selective,” and “indicated”
provide a valuable way to distinguish preventive interventions.
All 3 preventive intervention strategies refer to the different
target populations to which they are applied. Universal preventive
interventions are applied to whole populations. A childhood vaccination
program, a fluoride treatment of the water supply, programs to improve
dietary and exercise habits, and state-mandated health education
courses are all examples of universal preventive interventions.
Universal Prevention Programs
There are now several successful universal intervention programs—for
example, classroom, peer, school, and family interventions for children
as young as first grade—aimed at preventing conduct disorder
(CD) (5–7). These programs are designed to deliver interventions
to all children, even though some components may be delivered in
the classroom, some on the playground, and some at parent-family
meetings. Components focus on such areas as classroom management
of behaviour by the teacher, parental monitoring and supervision,
and positive peer relations, all known to protect against later
development of CD. One universal intervention strategy of potential
benefit in regard to schizophrenia would focus on lowering the incidence
of pregnancy and delivery complications through improved pre-, peri-,
and postnatal care.
Selective Prevention Programs
Conversely, selective interventions target specific subgroups.
Selective interventions target those who are at elevated risk, based
on group-level characteristics that are not directly related to
etiology. For example, children raised in families where there is
a recent divorce are at higher risk for internalizing and externalizing
problems (87). An intervention that provides parenting skills to
divorcing parents during this critical time is selective, because
it targets an entire subpopulation at risk, even though there are
certainly cases where some children fare much better after the divorce
(88). Other examples of selective interventions include a program
to prevent depression when an adult becomes unemployed (89), a family
program aimed at reducing depression in children whose parents have
a depressive disorder (2), and a drug-abuse prevention program for
children of heroin users (90).
An important challenge for schizophrenia prevention programs is
how to identify subpopulations for selective interventions (91).
A simple method would be to choose adolescent children or siblings
of schizophrenia patients. This group has a tenfold elevated risk
for the disorder and is entering the age period of greatest risk
for the onset of psychosis. However, even though the elevation in
risk for this group is substantial, only 10% can be expected to
develop schizophrenia or a related psychotic disorder. This magnitude
of risk is not sufficient to define the at-risk population for selective
prevention trials (unless there existed a low-risk treatment that
was inexpensive to administer).
Fortunately, research suggests that measures of schizotaxia may
some day improve risk prediction to the level where it would be
useful in defining populations at very high risk for schizophrenia
(18). For example, in 2 independent studies of children of schizophrenia
patients, Fish described a syndrome of motor abnormalities that
predicted subsequent schizophrenia or related disorders (92). Similarly,
in both the Copenhagen and New York high-risk projects, neuromotor
impairment predicted schizophrenia onset (93,94). These findings
are consistent with Walker and Lewine’s finding of poorer fine
and gross motor coordination in videotapes of children who subsequently
developed schizophrenia (95). In addition to neuromotor impairment,
attentional deficits have also been found to predict subsequent
schizophrenia and related disorders (for example, [94]).
Given the established link between neuropsychological and social
impairment in child relatives, it is not surprising that psychosocial
dysfunction also predicts subsequent schizophrenia and related disorders.
In the Israeli high-risk study, subjects who eventually developed
schizophrenia-related disorders had been either shy and withdrawn
or aggressive and antisocial as children (96). Walker and Lewine’s
videotape study described the children who developed schizophrenia
as having had poorer eye contact, more negative affect, and diminished
social responsiveness (95). Similarly, in the Copenhagen high-risk
study, teacher-rated social behaviours predicted subsequent schizophrenia
(93).
Thus, schizotaxic traits among children of schizophrenia patients
may predict subsequent psychosis. Yet, more needs to be known about
the diagnostic accuracy of schizotaxic traits (that is, their sensitivity
and specificity as predictors of psychosis [97]) before they can
be used in prevention trials. Many of the studies reviewed above
are difficult to interpret because the outcome they predict (that
is, schizophrenia and related disorders) is rather broad. Because
trials with antipsychotics would not be warranted for preventing
some related disorders (for example, schizotypal personality disorder
[SPD]), future work needs to estimate precisely the degree to which
schizotaxia can predict schizophrenia. Moreover, because no diagnostic
criteria are available for schizotaxia, it is not possible to compare
the diagnostic accuracy of schizotaxia and SPD as predictors of
psychosis. Future work will need to address this issue.
Theoretically, risk prediction should improve dramatically after
molecular genetic studies discover genes for schizophrenia. Notably,
linkage studies have discovered regions of the genome that may harbour
schizophrenia genes. Five promising chromosomal regions are 22q11-q13,
6p23, 8p22-21, 15q13-q14, and 10p14-p12 (98–107). However,
although these findings are promising, no schizophrenia gene has
yet been found. Once geneticists have identified the mutations leading
to schizophrenia, this information can be combined with schizotaxic
signs to delineate a group at very high risk for the disorder.
Indicated Prevention Programs
Indicated preventive interventions target individuals who either
have signs but are currently asymptomatic for a disorder or are
in an early stage of a progressive disorder. It should be noted
that in its original usage pertaining to heart disease, the term
“indicated intervention” referred only to those who were
asymptomatic; that is, those with borderline hypertension. Because
there are no universal signs of schizophrenia, indicated interventions
for this disorder have a somewhat broader definition than those
used in other health fields where clear signs are available (for
example, borderline hypertension for heart disease) (108).
The importance of indicated preventive interventions for schizophrenia
was suggested by Wyatt’s idea that early intervention for schizophrenia
might alter the course of the illness (109). His review of 21 studies
found that patients treated with antipsychotic medicine during their
first or second hospitalization had a better outcome than patients
treated later in the course of illness. Others have suggested that
early treatment, especially with newer agents, may preserve brain
plasticity and reduce the clinical deterioration occurring with
chronic schizophrenia (110,111). It is also possible that, rather
than having a neuroprotective effect, early treatment mitigates
the social consequences of schizophrenic psychopathology, which
may result in better outcome by allowing the easier reintegration
of patients into their social networks. This line of reasoning has
motivated the creation of early detection and indicated preventive
intervention projects seeking to treat schizophrenia patients during
their prodrome or first episode (93,110,112–117). Using signs
such as social withdrawal, subtle changes in thinking or affect,
and mild psychosis, these projects identify people in the very early
stages of schizophrenia.
For low base-rate disorders such as schizophrenia, one might conclude
that broad, population-based programs, and especially universal
programs, would not be as effective as targeted ones based on very
high-risk subgroups. After all, targeting a large group of subjects,
most of whom have very low risk for schizophrenia, seems inefficient,
compared with focusing more intensively on a smaller number of subjects
with higher risk. This inefficiency could occur when risk factors
identified in a high-risk group do not generalize to the population
For example, because children of parents with schizophrenia have
a high genetic loading for the disorder, it is possible that environmental
risk factors play a weaker role than they do in the general population.
Second, there are many situations wherein an intervention delivered
to everyone has greater effect than one given to a smaller group,
even though the smaller group has higher risk. Vaccinations, for
example, count on reducing the infectivity rate by reducing the
pool without inoculation. Similarly, behaviour-based interventions
can be highly effective when peers, classmates, or neighborhood
families participate together, rather than individually (7).
There is an important difference between indicated interventions
for schizophrenia and the selective interventions described in the
prior section. Whereas the indicated interventions target people
who have already begun to express signs and symptoms of the illness,
the selective interventions target people who only express the premorbid
condition known as schizotaxia. This difference is crucial: it is
possible that different interventions will be needed during the
premorbid and prodromal stages of illness. For example, reducing
a child’s exposure to stressful experiences may be effective
in childhood, during the premorbid stage. In late adolescence, by
contrast, reducing stress among families of prodromal patients may
modify the course of illness, but it cannot change the cumulative
impact of stress that occurred in earlier years.
1
| 2 | 3
| 4 | 5
|
