ORIGINAL ARTICLE

Evidence for Early-Childhood, Pan-Developmental

Impairment Specific to Schizophreniform Disorder
Results From a Longitudinal Birth Cohort
Mary Cannon, MD, PhD; Avshalom Caspi, PhD; Terrie E. Moffitt, PhD; HonaLee Harrington, BS;
Alan Taylor, MSc; Robin M. Murray, MD, DSc; Richie Poulton, PhD

Background: Childhood developmental abnormali- Results: Emotional problems and interpersonal diffi-
ties have been previously described in schizophrenia. It culties were noted in children who later fulfilled diag-
is not known, however, whether childhood developmen- nostic criteria for any of the adult psychiatric outcomes
tal impairment is specific to schizophrenia or is merely assessed. However, significant impairments in neuromo-
a marker for a range of psychiatric outcomes. tor, receptive language, and cognitive development were
additionally present only among children later diag-
Methods: A 1-year birth cohort (1972-1973) of 1037 chil- nosed as having schizophreniform disorder. Develop-
dren enrolled in the Dunedin Multidisciplinary Health and mental impairments also predicted self-reported psy-
Development Study was assessed at biennial intervals be- chotic symptoms at age 11 years. These impairments were
tween ages 3 and 11 years on emotional, behavioral, and independent of the effects of socioeconomic, obstetric,
interpersonal problems, motor and language develop- and maternal factors.
ment, and intelligence. At age 11 years, children were asked
about psychotic symptoms. At age 26 years, DSM-IV di- Conclusions: The results provide evidence for an early-
agnoses were made using the Diagnostic Interview Sched- childhood, persistent, pan-developmental impairment that
ule. Study members having schizophreniform disorder is specifically associated with schizophreniform disor-
(n=36 [3.7%]) were compared with healthy controls and der and that predicts psychotic symptoms in childhood
also with groups diagnosed as having mania (n=20 [2%]) and adulthood.
and nonpsychotic anxiety or depression disorders (n=278
[28.5%]) on childhood variables. Arch Gen Psychiatry. 2002;59:449-456

S
CHIZOPHRENIA IS a clinical syn- mation, follow-up studies of existing birth
drome with peak onset in late cohorts, and genetic high-risk studies that
adolescence or early adult- follow offspring of an affected parent
hood, whose symptoms are throughout childhood and adolescence.
manifest in multiple do- Such strategies have uncovered robust evi-
mains of behavior, language, thought, and dence for childhood motor, language, cog-
affect, and whose etiology remains ob- nitive, and behavioral precursors to schizo-
scure.1 A neurodevelopmental etiologic phrenia7-23 but there are 3 caveats. First,
model or hypothesis of schizophrenia has different developmental impairments have
From the Division of
been influential during the past decade.2,3 been examined in separate studies using
Psychological Medicine
(Drs Cannon and Murray), It proposes a subtle deviance in early brain a variety of case ascertainment methods
and the Social, Genetic, and development whose full adverse conse- and developmental scales. As a result, con-
Developmental Psychiatry quences do not emerge until adolescence clusions about the etiologic significance
Research Centre (Drs Caspi or early adulthood. Central to this hypoth- of developmental impairments may have
and Moffitt and Mr Taylor), esis is the identification of developmental been confounded by these variations.
Institute of Psychiatry, London, deficits preceding overt clinical symp- One should examine whether different
England; the University of toms of adult schizophrenia.4-6 In this study, types of developmental impairment pre-
Wisconsin–Madison, Madison we apply a life-course approach to the study dict the same adult schizophrenic out-
(Drs Caspi and Moffitt and
of schizophrenia and focus on develop- come in one longitudinal study. Second,
Ms Harrington); and the
Dunedin Multidisciplinary mental risk factors in early life evidence of specificity for schizophrenia
Health and Development Several different research strategies is limited. Childhood developmental
Research Unit, University of have been used to examine the develop- problems associated with schizophrenia
Otago, Dunedin, New Zealand mental precursors of adult schizophre- may also occur in patients with other
(Dr Poulton). nia, including the use of archived infor- psychiatric disorders24,25 and may thus be

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 59, MAY 2002 WWW.ARCHGENPSYCHIATRY.COM

449

©2002 American Medical Association. All rights reserved.

 PARTICIPANTS AND METHODS child; unaware or unresponsive to the child’s needs, indif-
ference to the child’s performance; demanding of the child’s
attention; and soiled, unkempt appearance of the child. This
SAMPLE assessment has been found in previous research to be re-
liable and valid.40 Scores on these 8 ratings were summed
Participants are members of the Dunedin Multidisciplinary to create a mother-child interaction variable for each mother,
Health and Development Study, a longitudinal investiga- and a score of 1 or more indicated atypical mother-child
tion of health and behavior in a complete birth cohort.35 The interaction for the purposes of this study.
study members were born in Dunedin, New Zealand, be-
tween April 1972 and March 1973. Of these, 1037 children Neuromotor Development
(91% of eligible births; 52% males) participated in the first
follow-up assessment at age 3 years, and they constitute the Infant milestones were assessed retrospectively at age 3 years.
base sample for the remainder of the study. Cohort families Mothers were asked to remember to the nearest month when
represent the full range of socioeconomic status (SES) in the their child attained various milestones: smiling, sitting up,
general population of New Zealand’s South Island and are walking, dry-by-day, dry-by-night, fed self, talked (words),
primarily white. Assessments have been conducted at ages and talked (sentences). Responses were recorded only when
3 (n=1037), 5 (n=991), 7 (n=954), 9 (n=955), 11 (n=925), the mother was certain that she could recall this informa-
13 (n=850), 15 (n=976), 18 (n=993), 21 (n=961), and most tion accurately. Most mothers referred to their “Plunkett
recently at age 26 years (n = 980; 96% of living cohort books,” in which study mothers had recorded this infor-
members). Participants are brought to the research unit within mation as their baby developed.
60 days of their birthday for a full day of individual data col- Neurologic abnormalities were assessed at age 3 years
lection. Various research topics are presented as standard- based on procedures described by Touwen and Prechtl.41
ized modules, each administered by a different trained ex- Each child was examined by a pediatric neurologist for neu-
aminer. Informed consent was obtained for all procedures. rologic signs, including assessment of motility, passive move-
ments, reflexes, facial musculature, strabismus, nystag-
CHILDHOOD MEASURES mus, foot posture, and gait. Motor development was assessed
at age 3 years with the Bayley Motor Scales,42 at age 5 years
SES, Obstetric Complications, and Maternal Factors using the McCarthy Motor Scales,43 and at ages 7 and 9 years
using the Basic Motor Ability Test.44
Family SES measured the average SES level of the study
members’ families across the first 15 years of life,36 using a Language and Cognitive Development
6-point scale designed for New Zealand where 1 = un-
skilled laborer and 6=professional.37 Receptive and expressive language development was as-
Each child was examined shortly after birth and pre- sessed at ages 3 and 5 years using the Reynell Developmen-
natal information was taken from the hospital records.38,39 tal Language Scales,45 which have separate subtests for re-
The obstetric complications assessed in this study were ma- ceptive (verbal comprehension) and expressive language. At
ternal diabetes; glycosuria; epilepsy; hypertension; eclamp- ages 7 and 9 years, language development was assessed us-
sia; antepartum hemorrhage; accidental hemorrhage; pla- ing the Auditory Reception and Verbal Expression subtests
centa previa; having had a previous small baby; gestational of the Illinois Test of Psycholinguistic Abilities.46
age younger than 37 weeks or older than 41 weeks; birth Intelligence was assessed at age 3 years with the Pea-
weight less than 2500 g or greater than 4 kg; small or large body Picture Vocabulary Test,47 at age 5 years with the Stan-
for gestational age; major or minor neurologic signs; Rh ford-Binet Intelligence Scales,48 and at ages 7, 9, and 11 years
incompatibility; ABO incompatibility; nonhemolytic hy- with the Weschler Intelligence Scales for Children–
perbilirubinemia; hypoxia at birth (idiopathic respiratory Revised.49 All tests were administered by trained psychom-
distress syndrome or apnea), and low Apgar score at birth. etrists according to standard protocol.50
(The infant was defined as having a low Apgar score if one
of the following conditions applied: [1] at 5 minutes of life, Internalizing and Externalizing Behavior Problems
the infant’s heart rate was ⬍100 beats/min, respiration was
irregular or absent, and the infant was centrally cyanosed; At ages 5, 7, 9, and 11 years, parents and teachers com-
[2] the infant took more than 10 minutes to establish nor- pleted the Rutter Child Scales,51,52 which inquire about chil-
mal respiration; or [3] the infant’s asphyxia at birth war- dren’s emotional and behavioral functioning during the past
ranted resuscitation.) Each complication was weighted year. The Internalizing Problems scale describes children
equally and summed to yield an obstetric complication who worry about many things or who often appear miser-
index.38 able, unhappy, and tearful. The Externalizing Problems scale
Mothers were rated on their general attitude and be- describes children who frequently fight, bully other chil-
havior in relation to their child by a psychologist or medi- dren, lie, steal, disobey, truant, destroy property, and have
cal doctor during the course of the child’s assessment at irritable tempers. The relevant items were summed across
age 3 years. Mothers were rated on 8 features: harshness the 4 age periods and 2 raters (parents and teachers) to de-
toward the child; critical or negative evaluation of the child; rive measures indexing children’s internalizing and exter-
rough, awkward handling of the child; no effort to help the nalizing problems, respectively.

nonspecific markers for a wide range of psychologic cause is suggested by the occurrence of developmental
disturbances in adulthood. A third area of debate is the problems in 25% to 40% of children at genetically high
etiology of such developmental precursors. A genetic risk for schizophrenia7,8,16 but it has been suggested that

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 59, MAY 2002 WWW.ARCHGENPSYCHIATRY.COM

450

©2002 American Medical Association. All rights reserved.

 Interpersonal Adjustment STATISTICAL ANALYSIS

At ages 5, 7, 9, and 11 years, parents evaluated 2 state- Analyses compare 4 mutually exclusive groups defined ac-
ments about their child: “my child is a loner” and “my child cording to psychiatric outcomes at age 26 years: schizo-
is not much liked by other children.” Each statement was phreniform disorder, mania, anxiety/depression, and a con-
rated on a 3-point scale. At ages 7, 9, and 11 years, teach- trol group composed of the remainder of the cohort, who
ers independently evaluated the same statements. Mean had none of the aforementioned disorders. ␹2 Tests were
scores for each statement were calculated separately for par- used to examine the associations among adult psychiatric
ents and teachers and these ratings were averaged for each disorders, sex, and family SES.
child to derive 2 measures indexing social isolation and peer The raw scores for childhood developmental vari-
rejection, respectively. ables were standardized, within age, on the entire cohort,
using the z-score transformation so that the cohort had a
Psychotic Symptoms at Age 11 Years mean of 0 and an SD of 1 on these variables. The figures
show the standardized scores for each outcome group. Dif-
At age 11 years, 789 study members were administered the ferences between outcome groups can be evaluated by com-
Diagnostic Interview Schedule for Children53 by a child psy- paring differences in z scores (SD units), where 0.2 is a small,
chiatrist.54 The schizophrenia section of the Diagnostic In- 0.5 is a moderate, and 0.8 is a large effect size.57
terview Schedule for Children was composed of 5 ques- Relationships between childhood developmental im-
tions regarding possible psychotic symptoms, which were pairments and psychiatric outcomes at age 26 years were ex-
scored by the psychiatrist as no (0); yes, likely (1); and yes, amined using a collection of regression techniques as re-
definitely (2). The scores for each item were summed. Most quired by the different types of developmental variables
study members (n=673) obtained a score of 0, 103 (13%) examined in this study (categorical, continuous, and
obtained a score of 1 and were called the weak-symptom repeated). Each regression equation was composed of 3
group, and the remaining 13 obtained a score of 2 or higher dummy variables for diagnostic status (schizophreniform,
and were called the strong-symptom group. Individuals in mania, and anxiety/depression groups), with the control
the strong-symptom group at age 11 years were found to group as the reference category. All reported regression co-
have a very high risk of schizophreniform disorder at age efficients and ORs were adjusted for sex and SES. Logistic
26 years (odds ratio [OR], 16.4; 95% confidence interval regression analysis examined the following categorical vari-
[CI], 3.9-67.8).34 Individuals in the weak-symptom group ables: individual obstetric complications, maternal rejec-
also had an increased risk of schizophreniform disorder at tion, and presence of 1 or more neurologic signs at age 3 years.
age 26 years but to a lesser degree (OR, 5.1; 95% CI, Ordinary least squares regression examined the following con-
1.7-18.3). tinuous variables: peer rejection and social isolation. Motor
and language development, IQ, and internalizing and ex-
Psychiatric Status at Age 26 Years ternalizing problems were measured on multiple occasions
and analyzed using the generalized estimating equation (GEE)
Psychiatric interviews at age 26 years were available for 976 approach—a form of repeated-measures regression analy-
of the 1019 cohort members still living. The Diagnostic In- sis in which any required covariance structure may be as-
terview Schedule55 was administered by health profession- sumed and parameters estimated without specifying the joint
als with either a medical or master’s degree to yield DSM-IV distribution of the repeated observations.58 We specified an
diagnoses.56 The reporting period was 12 months prior to the unstructured correlation matrix and used robust SEs to pro-
interview. The Axis I disorders diagnosed at age 26 years were tect against model misspecification.59 The GEE approach can
grouped into the following diagnostic outcome groups: (1) accommodate noninformative missing values.60,61 We re-
schizophreniform disorder (n=36 [3.7%]), (2) manic epi- port regression coefficients adjusted for sex and SES and their
sodes (n=20 [2.0%]), and (3) anxiety or depressive disor- 95% CIs. These coefficients represent the average differ-
ders (n=278 [28.5%]). The primary outcome for this study ence among diagnostic groups. To test whether relation-
was schizophreniform disorder. Diagnostic procedures for ships between developmental impairments and psychiatric
schizophreniform disorder are explained in detail by Poul- outcomes at age 26 years were obtained independently of
ton et al.34 To enhance the validity of our research diagnosis, perinatal and postnatal environmental factors, all GEE analy-
we took 2 additional steps: (1) We required the presence of ses were repeated, controlling for obstetric complications and
hallucinations (not substance-related) plus at least 2 other maternal rejection.
symptoms from Criterion A of the DSM-IV (this is more strict To test whether the developmental impairments that
than the DSM-IV diagnostic criteria). and (2) We required were associated with a schizophreniform outcome at age
objective evidence of impairment from informants to comple- 26 years were also associated with psychotic symptoms at
ment self-reports. Following this protocol, 1% of the sample age 11 years, we repeated the GEE analyses using 2 dummy
met criteria for formal schizophrenia at age 26 years and a variables representing the weak- and strong-symptom groups
further 2.7% met all criteria except 6-month chronicity. For at age 11 years, with the nonsymptom group as the refer-
the purposes of this analysis, study members who were co- ence category. All analyses were carried out using Stata ver-
morbid for 2 or more disorders were assigned to 1 of 3 diag- sion 6.0 (Stata Corp, College Station, Tex).62 Interactions
nostic groups, in the following order of priority: schizophreni- between sex and diagnosis were not examined owing to
form disorder, mania, and anxiety/depression. power limitations. All significance tests were 2-tailed.

maternal and social factors26-29 or obstetric factors7,8,30 processes involved in the genesis of psychopathology
may be partly responsible for these associations. There- should also incorporate perinatal and postnatal envi-
fore, a comprehensive investigation of developmental ronmental factors.31-33

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 59, MAY 2002 WWW.ARCHGENPSYCHIATRY.COM

451

©2002 American Medical Association. All rights reserved.

 0.8 sociation with later mania (␤=.03; 95% CI, −0.15 to 0.21)
Control or anxiety/depression (␤=−.03; 95% CI, −0.08 to 0.03). Post
0.6 Anxiety/Depression
Motor Test Performance, z Scores Mania hoc analyses revealed that 3 complications were associ-
0.4 Schizophreniform ated with an increased risk of schizophreniform disorder:
low Apgar score at birth (OR, 5.9; 95% CI, 1.1-32.0); hy-
0.2
poxia at birth (apnea or idiopathic respiratory distress syn-
0 drome) (OR, 5.0; 95% CI, 1.5-16.4); and small-for-
–0.2
gestational-age status (OR, 2.8; 95% CI, 1.1-6.9). The
mothers of the schizophreniform group (OR, 2.65; 95% CI,
–0.4
1.2-5.6) but not the manic group (OR, 1.7; 95% CI, 0.6-
–0.6 4.9) or the anxiety/depression group (OR, 1.4; 95% CI, 0.9-
2.03) were significantly more likely to have atypical mother-
–0.8
3 5 7 9 child interactions when compared with mothers of controls.
Age, y
NEUROMOTOR DEVELOPMENT
Figure 1. The mean standardized scores for motor development at ages 3, 5,
7, and 9 years for adults diagnosed as having schizophreniform disorder AND PSYCHIATRIC OUTCOMES
(n = 36), mania (n = 20), anxiety/depression (n=278), and controls (n = 642).
The schizophreniform group began to walk significantly
later than controls (mean [SE], 14.9 [1.0] months vs 13.6
The longitudinal Dunedin Study has followed up 1000 [0.13]months;F3,661=5.01[adjustedforsexandSES];P=.02)
children from the general population from birth to age 26 but there were no differences for any other infant milestones.
years and offers several advantages for investigating child- At age 3 years, the schizophreniform group was significantly
hood risk factors for adult outcomes. First, we can exam- more likely than controls to have one or more neurologic
ine multiple developmental risk factors within one sample signs (OR, 4.6; 95% CI, 1.9-10.9). The mania group (OR,
using detailed childhood data that were prospectively col- 0.8; 95% CI, 0.1-6.4) and the anxiety/depression group (OR,
lected from age 3 years using well-established and vali- 1.7; 95% CI, 0.9-2.8) were not significantly more likely to
dated instruments. Second, we can examine the specific- have neurologic signs than controls. The schizophreniform
ity of early developmental risk factors for schizophreniform group also performed worse than controls (more than 0.3
outcomes since psychiatric diagnoses were made for all SDs) on standard tests of motor skill at ages 3, 5, and 9 years
study members at age 26 years based on structured diag- but not at age 7 years (Figure 1). The repeated-measures
nostic interviews conducted by trained health profession- analysisshowedthattheschizophreniformgroupperformed
als. A third unique strength is that we can study whether significantly worse than the control group overall, while the
the same developmental risk factors predict psychotic symp- mania group performed significantly better than the con-
toms in childhood and in adulthood. In an earlier report trol group on motor performance, even after controlling for
from the Dunedin cohort, we showed that children’s self- sex and SES (Table 1). The anxiety/depression group did
reported psychotic symptoms at age 11 years predicted a not differ significantly from controls on any of these mo-
schizophreniform diagnosis at age 26 years.34 If the same tor assessments.
(or similar) relationships are found between childhood de-
velopmental risk factors and psychotic symptoms at age 11 LANGUAGE DEVELOPMENT, COGNITIVE
years as with schizophreniform disorder at age 26 years, it DEVELOPMENT, AND PSYCHIATRIC OUTCOMES
would suggest that the psychotic symptoms at age 11 years
are part of the disease process itself rather than an inde- The schizophreniform group did not exhibit any prob-
pendent risk factor/marker for later schizophreniform dis- lems with expressive language but their receptive lan-
order. guage skills were significantly poorer than those of the
controls (between 0.2-0.6 SDs) at each of the 4 biennial
RESULTS testings during the first decade of life (Figure 2). The
schizophreniform group also performed more poorly than
There were significant overall sex and family SES differ- controls (about 0.4 SDs) on standard IQ tests at each of
ences among the adult diagnostic groups. The adult anxi- 5 assessments between ages 3 and 11 years (Figure 3).
ety/depression group contained significantly more fe- These significant impairments in receptive language and
males than the control group (60.1% vs 45%; ␹21 =17.6; cognitive development among the schizophreniform
P⬍.01) and a significantly higher proportion of adults group were independent of sex and SES (Table 1). The
in the schizophreniform group came from low-SES fami- mania and the anxiety/depression groups did not differ
lies (categories 1 and 2) compared with controls (47.2% significantly from controls on either language measure
vs 9.2%; ␹21 =16.5; P⬍.01). or on IQ test performance (Table 1).

OBSTETRIC COMPLICATIONS, MATERNAL INTERNALIZING AND EXTERNALIZING

FACTORS, AND PSYCHIATRIC OUTCOMES PROBLEMS, INTERPERSONAL ADJUSTMENT,
AND PSYCHIATRIC OUTCOMES
There was a significant association between the obstetric
complications index and later schizophreniform disorder The schizophreniform group and the anxiety/depression
(␤=.38; 95% CI, 0.25-0.52; P=.02) but no significant as- group exhibited significantly more childhood internaliz-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 59, MAY 2002 WWW.ARCHGENPSYCHIATRY.COM

452

©2002 American Medical Association. All rights reserved.

 Table 1. Results From Regression Analyses Showing the Association Between Developmental Functioning
During the First Decade of Life and Adult Psychiatric Outcomes at Age 26 Years*

Psychiatric Outcomes at Age 26 Years

Schizophreniform Disorder Mania Anxiety/Depression

(n = 36) (n = 20) (n = 278)

Developmental Variables Coefficient (95% CI) P Value Coefficient (95% CI) P Value Coefficient (95% CI) P Value
Motor development, age 3-9 y −0.39 (−0.67 to −0.12) .005 0.33 (0.06-0.6) .01 −0.05 (−0.15 to 0.04) .3
Receptive language, age 3-9 y −0.31 (−0.54 to −0.06) .01 0.02 (−0.2 to 0.24) .8 −0.04 (−0.14 to 0.06) .4
Expressive language, age 3-9 y 0.09 (−0.15 to 0.34) .4 0.11 (−0.16 to 0.37) .4 −0.03 (−0.12 to 0.06) .5
IQ, age 3-11 y −0.33 (−0.6 to −0.07) .01 0.11 (−0.1 to 0.32) .3 −0.07 (−0.19 to 0.03) .16
Internalizing problems, age 5-11 y 0.27 (0.02-0.53) .03 0.28 (−0.01 to 0.58) .06 0.17 (0.05-0.28) .004
Externalizing problems, age 5-11 y 0.32 (−0.05 to 0.7) .09 0.36 (−0.04 to 0.75) .07 0.09 (−0.01 to 0.2) .09
Peer rejection, mean score, age 5-11 y 0.46 (0.01-0.9) .04 0.54 (0.02-1.06) .04 0.16 (0.01-0.31) .03
Social isolation, mean score, age 5-11 y 0.17 (−0.18 to 0.53) .3 0.46 (0.09-0.82) .01 0.16 (0.01-0.31) .04

*The reference group in the regression equations comprises control subjects who did not meet diagnostic criteria for schizophreniform disorder, mania,
or anxiety/depression at age 26 years (n = 642). The regression coefficients are interpretable as SD unit differences between each psychiatric group
and the control group, adjusted for sex and family socioeconomic status. Negative coefficients for motor development, receptive language, expressive language,
and IQ (intelligence quotient) indicate worse performance; positive coefficients for internalizing problems, externalizing problems, peer rejection, and
social isolation indicate worse adjustment. CI indicates confidence interval.

ing problems as rated by parents and teachers than the con- A Expressive Language B Receptive Language
trol group (Figure 4). These effects were independent 0.6
of sex and SES (Table 1). All 3 diagnostic groups exhib- Control

Language Test Performance, z Scores

Anxiety/Depression
ited more childhood externalizing problems than the con- 0.4
Mania
trol group (Figure 4) but, when adjusted for sex and SES, Schizophreniform
0.2
these differences just missed significance at the 5% level
(Table 1). All 3 diagnostic groups were significantly more 0
likely than the control group to be rejected by peers, as
rated by parents and teachers (Table 1). The mania group –0.2
and the anxiety/depression group, but not the schizo-
phreniform group, were significantly more likely than con- –0.4

trols to be rated by parents and teachers as socially iso-

–0.6
lated (Table 1). 3 5 7 9 3 5 7 9
Age, y Age, y
OBSTETRIC AND MATERNAL FACTORS IN Figure 2. The mean standardized scores for expressive and receptive
RELATION TO DEVELOPMENTAL IMPAIRMENT language development at ages 3, 5, 7, and 9 years for adults diagnosed as
having schizophreniform disorder (n = 36), mania (n = 20), anxiety/depression
We investigated whether the relationships between de- (n = 278), and controls (n = 642).
velopmental deficits and schizophreniform disorder were
independent of obstetric complications and atypical 0.4
Control
mother-child interaction. Adjusting for the obstetric com- Anxiety/Depression
plications that were significantly related to schizophreni- 0.2 Mania
IQ Test Performance, z Scores

form outcome and for atypical mother-child interaction Schizophreniform

(as well as sex and SES), the associations between schizo-

0
phreniform disorder and motor development (␤= −.35;
95% CI, −0.59 to −0.12; P = .003), receptive language
(␤=−.22; 95% CI, –0.45 to 0.02; P=.07), and IQ (␤=−.26; –0.2

95% CI, –0.52 to −0.004; P = .04) did not change signifi-

cantly. –0.4

CHILDHOOD DEVELOPMENTAL IMPAIRMENT –0.6

AND PSYCHOTIC SYMPTOMS AT AGE 11 YEARS 3 5 7 9 11
Age, y

Self-reported strong psychotic symptoms at age 11 years Figure 3. The mean standardized scores for intelligence tests at ages 3, 5, 7,
were associated with significant developmental impair- 9, and 11 years for adults diagnosed as having schizophreniform disorder
ments in neuromotor development, receptive language, (n = 36), mania (n = 20), anxiety/depression (n = 278), and controls (n=642).
intelligence, and emotional development (Table 2). The
effect sizes were generally even larger than those noted not significantly associated with childhood developmen-
for schizophreniform disorder at age 26 years. Apart from tal impairments. However, the direction of the coeffi-
an association with receptive language impairment, self- cients for the weak-symptom group was in the same di-
reported weak psychotic symptoms at age 11 years were rection as the coefficients for the strong-symptom group.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 59, MAY 2002 WWW.ARCHGENPSYCHIATRY.COM

453

©2002 American Medical Association. All rights reserved.

 COMMENT ness processes that begin in childhood, and that child-
hood psychotic symptoms may be part of a disease pro-
This longitudinal investigation of an unselected birth co- cess rather than an independent risk factor/marker for
hort examined several childhood risk factors in relation later schizophreniform disorder.
to 3 adult psychiatric outcomes. Children who later ful- The relationship between neuromotor developmen-
filled diagnostic criteria for schizophreniform disorder tal problems and later schizophreniform disorder in this
at age 26 years exhibited significant impairments across study was particularly strong, with evidence of delay in
a range of developmental domains (neuromotor, lan- learning to walk during infancy, an excess of neurologic
guage, cognitive, emotional, and interpersonal develop- signs at age 3 years, and significant impairments on re-
ment) from as young as 3 years. In contrast, children who peated motor testing between ages 3 and 9 years. We found
later fulfilled diagnostic criteria for mania and anxiety/ that the schizophreniform group exhibited deficits in re-
depression exhibited problems only in the areas of emo- ceptive language development (verbal comprehension)
tional and interpersonal development. Early neuromo- rather than expressive language development. Previous
tor, language, and cognitive developmental impairments work on this cohort has shown that receptive but not ex-
therefore seem to show specificity to schizophreniform pressive language delay at age 3 years was significantly
disorder, whereas childhood emotional and interper- associated with behavior problems in late childhood,63
sonal difficulties are associated with a range of psychi- and a follow-up study of a group of children with devel-
atric disorders in adulthood. Of special interest is the find- opmental receptive language disorder has found that 10%
ing that similar childhood developmental deficits were of the children later developed schizophrenia.64 Our study
observed in relation to self-reported psychotic symp- also revealed that childhood cognitive impairments among
toms at age 11 years and in relation to schizophreni- the schizophreniform group were emerged early and were
form disorder at age 26 years. This suggests that these persistent, with significant deficits in IQ detectable from
developmental deficits are associated with psychotic ill- age 3 years. In sum, our findings on motor, language, and
cognitive impairments add to a body of work showing
that childhood developmental deficits are found among
A Internalizing Problems B Externalizing Problems
0.8
individuals with schizophrenia and among those at ge-
Control netic risk for schizophrenia.7-23 Our study provides new
Anxiety/Depression evidence that these impairments may be specific to schizo-
Child Behavior Ratings, z Scores

0.6
Mania
Schizophreniform
phrenia. Although 2 previous studies have noted some
0.4 (albeit weaker) associations between childhood motor and
speech problems and affective disorder, these abnormali-
0.2
ties were mainly confined to childhood-onset cases.24,25
0
Recent work has found that obstetric complica-
tions involving hypoxia and fetal growth retardation are
–0.2 risk factors for schizophrenia,65-69 and such effects were
also noted in this study. In agreement with other cohort
–0.4
5 7 9 11 5 7 9 11
studies,9,70 we found that aspects of the mother-child in-
Age, y Age, y teraction were associated with later schizophreniform dis-
order. However, these perinatal and maternal risk fac-
Figure 4. The mean standardized scores for internalizing and externalizing
problems at ages 5, 7, 9, and 11 years for adults diagnosed as
tors could not entirely account for the early developmental
havingschizophreniform disorder (n=36), mania (n=20), anxiety/depression impairments found in our schizophreniform group. We
(n = 278), and controls (n=642). therefore surmise, along with others,20-22 that neurode-

Table 2. Results From Regression Analyses Showing the Association Between Developmental
Functioning During the First Decade of Life and Psychotic Symptoms at Age 11 Years*

Psychotic Symptoms at Age 11 y

Weak Symptoms (Score = 1) (n = 103) Strong Symptoms (Score⬎1) (n=13)

Developmental Variables Coefficient (95% CI) P Value Coefficient (95% CI) P Value
Motor development, age 3-9 y −0.11 (−0.25 to −0.02) .11 −0.62 (−1.2 to −0.04) .03
Expressive language, age 3-9 y 0.015 (−0.11 to 0.14) .8 −0.11 (−0.42 to 0.19) .5
Receptive language, age 3-9 y −0.14 (−0.27 to −0.005) .04 −0.57 (−0.92 to −0.22) ⬍.01
IQ, age 3-11 y −0.12 (−0.26 to 0.03) .13 −0.52 (−0.83 to −0.21) ⬍.01
Internalizing problems, age 5-11 y 0.15 (−0.05 to 0.34) .13 0.74 (0.18 to 1.3) .01
Externalizing problems, age 5-11 y 0.15 (−0.03 to 0.33) .11 0.34 (−0.28 to 0.94) .28
Peer rejection, mean score, age 5-11 y 0.17 (−0.03 to 0.37) .10 0.51 (−0.2 to 1.2) .16
Social isolation, mean score, age 5-11 y 0.03 (−0.17 to 0.24) .73 0.46 (−0.14 to 1.06) .14

*The reference group in the regression equations includes subjects who did not report any psychotic symptoms at age 11 years (n = 673). The regression
coefficients are interpretable as SD unit differences between the weak- and strong-symptom groups and the control group, adjusted for sex and family
socioeconomic status. Negative coefficients for motor development, receptive language, expressive language, and IQ (intelligence quotient) indicate worse
performance; positive coefficients for internalizing problems, externalizing problems, peer rejection, and social isolation indicate worse adjustment. CI indicates
confidence interval.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 59, MAY 2002 WWW.ARCHGENPSYCHIATRY.COM

454

©2002 American Medical Association. All rights reserved.

velopmental impairments are not merely mediators of the Submitted for publication February 5, 2001; final revision
effects of obstetric complications on risk for schizophre- received August 6, 2001; accepted September 4, 2001.
nia. These early developmental impairments are more likely Dr Cannon was supported by an Advanced Fellowship
to reflect the expression of schizophrenia-susceptibility from the Wellcome Trust, London, England, and an EJLB
genes,71,72 and reports of developmental impairments Scholar Research Award from the EJLB Foundation, Mont-
among offspring at high genetic risk for schizophrenia real, Quebec. Additional funding was provided by the Schizo-
lend support to this view.7,8,16,19 phrenia Research Fund, London, England (Drs Cannon, Caspi
Emotional problems and poor interpersonal func- and Moffitt). The Dunedin Multidisciplinary Health and De-
tioning in childhood were associated with a host of dif- velopment Study is supported by the Health Research Coun-
ferent adult psychiatric outcomes at age 26 years, includ- cil of New Zealand, Auckland, and grants MH49414 (Dr
ing schizophreniform disorder, manic episodes, and Caspi), MH45548, and MH45070 (Dr Moffitt) from the Na-
anxiety/depression disorders. These predictive associa- tional Institute of Mental Health, Bethesda, Md.
tions are of actuarial interest—they span more than 15 We thank Phil Silva, PhD, founding director of the study,
years and do not exhibit specificity to any one outcome. Air New Zealand, the study members and their families for
Lack of specificity is important because it indicates a com- their participation and support, Matt Smart for secretarial
mon pathway to the development of a range of different assistance, and Professor Sir Michael Rutter, FRS, and Pro-
disorders. Although this constellation of childhood be- fessor Peter Jones, PhD, for comments on the manuscript.
haviors observed in children as young as 5 years is un- Corresponding author and reprints: Terrie E. Moffitt,
likely to represent a prodrome, it may index, more gen- PhD, Social, Genetic, and Developmental Psychiatry Re-
erally, a vulnerable personality that is at risk for all adult search Centre, Institute of Psychiatry, Box P080, London
psychiatric disorders. Although we found no significant SE5 8AF, England (e-mail: t.moffitt@iop.kcl.ac.uk).
association between childhood social isolation and schizo-
phreniform disorder, as noted by others,9,73-75 it is pos-
sible that such peer problems become more evident dur- REFERENCES
ing adolescence.76,77
1. Andreasen NC. Understanding the causes of schizophrenia. N Engl J Med. 1999;
Limitations of our study should be noted. First, the 340:645-647.
sample sizes in the schizophreniform and manic groups 2. Weinberger DR. Implications of normal brain development for the pathogenesis
are not large. Nevertheless, the developmental impair- of schizophrenia. Arch Gen Psychiatry. 1987;44:660-669.
3. Murray RM, & Lewis SW. Is schizophrenia a neurodevelopmental disorder? BMJ.
ments in the schizophreniform group were consistently 1987;295:681-682.
detected on repeated testings throughout childhood and 4. Jones PB. Longitudinal approaches to the search for the causes of schizophre-
were robust to adjustment for sex and social class nia: past, present and future. In: Search for the Causes of Schizophrenia. Vol 4.
Berlin, Germany: Springer; 1999.
effects. Second, the study members have not passed 5. Marenco S, Weinberger DR. The neurodevelopmental hypothesis of schizophre-
through the entire period of risk for psychosis. How- nia: following a trail of evidence from cradle to grave. Dev Psychopathol. 2000;
12:501-527.
ever, the childhood developmental risk factors found in 6. McDonald C, Fearon P, Murray RM. Neurodevelopmental hypothesis of schizo-
this study are remarkably similar to those found in a phrenia 12 years on: data and doubts. In: Rapoport JL, ed. Childhood Onset of
cohort study that has followed up participants to age 43 Adult Psychopathology. Washington, DC: American Psychiatric Press; 2000:
193-222.
years,9 suggesting that our findings can be extrapolated 7. Fish B, Marcus J, Hans SL, Auerbach JG, Perdue S. Infants at risk for schizo-
throughout the age-incidence distribution. Last, phrenia: sequelae of a genetic neurointegrative defect: a review and replication
throughout this analysis we have reported on findings analysis of pandysmaturation in the Jerusalem Infant Development Study. Arch
Gen Psychiatry. 1992;49:221-235.
for schizophreniform disorder rather than schizophre- 8. Marcus J, Hans SL, Auerbach JG, Auerbach AG. Children at risk for schizophre-
nia alone, partly because of sample-size considerations nia: the Jerusalem Infant Development Study II: neurobehavioural deficits at school
age. Arch Gen Psychiatry. 1993;50:797-809.
and also because dimensional or continuum models of 9. Jones P, Rodgers B, Murray R, Marmot, M. Child developmental risk factors for
psychosis are becoming established as the most likely adult schizophrenia in the British 1946 birth cohort. Lancet. 1994;344:1398-
theoretically78 and the most useful clinically.79 Indeed, it 1402.
10. Done DJ, Crow TJ, Johnstone EC, Sacker A. Childhood antecedents of schizo-
is impressive that the same childhood developmental phrenia and affective illness: social adjustment at ages 7 and 11. BMJ. 1994;
risk factors seem to apply to the broader phenotype of 309:699-703.
schizophreniform disorder as to the narrower concept 11. Walker E, Savoie T, Davis D. Neuromotor precursors of schizophrenia. Schizophr
Bull. 1994;20:441-451.
of schizophrenia itself. 12. Hollis C. Child and adolescent (juvenile onset) schizophrenia: a case control study
In conclusion, this study demonstrates that schizo- of premorbid developmental impairments. Br J Psychiatry. 1995;166:489-495.
13. Crow TJ, Done DJ, Sacker A. Childhood precursors of psychosis as clues to its
phreniform disorder is associated with childhood devel- evolutionary origins. Eur Arch Psychiatry Clin Neurosci. 1995;245:61-69.
opmental deficits across a range of domains. Motor, lan- 14. Olin SS, Mednick SA. Risk factors of psychosis: identifying vulnerable popula-
guage, and cognitive developmental deficits emerge early tions premorbidly. Schizophr Bull. 1996;22:223-240.
15. Kremen WS, Buka SL, Seidman LJ, Goldstein JM, Koren D, Tsuang MT. IQ de-
and are persistent and specific to schizophreniform cline during childhood and adult psychotic symptoms in a community sample: a
disorder. In addition, we have shown that pan-de- 19-year longitudinal study. Am J Psychiatry. 1998;155:672-677.
velopmental impairments are associated with psychotic 16. Hans SL, Marcus J, Nuechterlain KH, Asarnow RF, Styr B, Auerbach JG. Neu-
robehavioural deficits at adolescence in children at risk for schizophrenia: the
symptoms both in childhood and in adulthood. Taken Jerusalem Infant Development Study. Arch Gen Psychiatry. 1999;56:741-748.
as a whole, the evidence from this study provides sup- 17. Amminger GP, Pape S, Rock D, Roberts SA, Ott SL, Squires-Wheeler E, Kesten-
baum C, Erlenmeyer-Kimling L. Relationship between childhood behavioural dis-
port for a neurodevelopmental model of schizophre- turbance and later schizophrenia in the New York High Risk Project. Am J Psy-
nia,2-6 echoing the earlier theoretical concept of schizo- chiatry. 1999;156:525-530.
taxia.80-82 It is increasingly evident that understanding the 18. Cannon M, Jones P, Huttunen MO, Tanskanen A, Huttunen T, Rabe-Hesketh S,
Murray RM. School performance in Finnish children and later development of
complex mechanisms governing brain development will schizophrenia: a population-based longitudinal study. Arch Gen Psychiatry. 1999;
ultimately hold the key to the etiology of schizophrenia. 56:457-463.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 59, MAY 2002 WWW.ARCHGENPSYCHIATRY.COM

455

©2002 American Medical Association. All rights reserved.

19. Erlenmeyer-Kimling L, Rock D, Roberts SA, Janal M, Kestenbaum C, Cornblatt 50. Moffitt TE, Caspi A, Harkness AR, Silva PA. The natural history of change in in-
B, Adamo UH, Gottesman II. Attention, memory and motor skills as childhood tellectual performance: who changes? how much? is it meaningful? J Child Psy-
predictors of schizophrenia-related psychoses: the New York High Risk Project. chol Psychiatry. 1993;34:455-506.
Am J Psychiatry. 2000;157:1416-1422. 51. Rutter M, Tizard J, Whitmore K. Education, Health and Behaviour. London, En-
20. Rosso IM, Bearden CE, Hollister JM, Gasperoni TL, Sanchez LE, Hadley T, Can- gland: Longmans; 1970.
non TD. Childhood neuromotor dysfunction in schizophrenia patients and their 52. Elander J, Rutter M. Use and development of the Rutter parents’ and teachers’
unaffected siblings: a prospective cohort study. Schizophr Bull. 2000;26:367- scales. Int J Methods Psychiatr Res. 1996;6:63-78.
378. 53. Costello A, Edelbrook C, Kalas R, Kessler M, Klaric S. NIMH Diagnostic Inter-
21. Bearden CE, Rosso IM, Hollister JM, Sanchez LE, Hadley T, Cannon TD. A pro- view for Children: Child Version. Rockville, Md: National Institute of Mental Health;
spective cohort study of childhood behavioural deviance and language abnor- 1982.
malities as predictors of adult schizophrenia. Schizophr Bull. 1999;26:395-410. 54. Anderson JC, Williams S, McGee R, Silva PA. DSM-III disorders in preadoles-
22. Cannon TD, Bearden CE, Hollister JM, Rosso IM, Sanchez LE, Hadley T. Child- cent children: prevalence in a large sample from the general population. Arch
hood cognitive functioning in schizophrenia patients and their unaffected sib- Gen Psychiatry. 1987;44:69-74.
lings: a prospective cohort study. Schizophr Bull. 2000;26:379-393. 55. Robins LN, Cottler L, Bucholz K, Compton W. Diagnostic Interview Schedule for
23. Isohanni M, Jones PB, Kemppainen L, Croudace T, Isohanni I. Childhood and DSM-IV. St Louis, Mo: Washington University; 1995.
adolescent precursors of schizophrenia in the Northern Finland 1966 Birth Cohort: a 56. American Psychiatric Association. Diagnostic and Statistical Manual of Mental
descriptive life-span model. Eur Arch Psychiatry Clin Neurosci. 2000; Disorders, Fourth Edition. Washington, DC: American Psychiatric Association;
250:1-9. 1994.
24. Van Os J, Jones P, Lewis G, Wadsworth M, Murray R. Developmental precur- 57. Cohen J. A power primer. Psychol Bull. 1992;112:155-159.
sors of affective illness in a general population birth cohort. Arch Gen Psychia- 58. Liang KY, Zeger SL. Longitudinal data analysis using generalised linear models.
try. 1997;54:625-631. Biometrika. 1986;73:13-22.
25. Sigurdsson E, Fombonne E, Sayal K, Checkley S. Neurodevelopmental anteced- 59. Rabe-Hesketh S, Everitt BS. A Handbook of Statistical Analyses Using Stata. Boca
ents of early-onset bipolar disorder. Br J Psychiatry. 1999;174:121-127. Raton, Fla: Chapman & Hall/CRC Press; 2000.
26. Sobel D. Children of schizophrenic patients: preliminary observations on early 60. Diggle PJ, Liang KY, Zeger SL. Analysis of Longitudinal Data. Oxford, England:
development. Am J Psychiatry. 1961;118:512-517. Oxford University Press; 1994.
27. Ragins N, Schachter J, Elmer E, Preisman R, Bowes AE, Harway V. Infants and 61. Everitt BS. Analysis of longitudinal data: beyond MANOVA. Br J Psychiatry. 1998;
children at risk for schizophrenia: environmental and developmental observa- 172:7-10.
tions. Am J Child Psychiatry. 1975;14:150-177. 62. StataCorp. Intercooled Stata 6.0 for Windows 98/95/NT. College Station, Tex:
28. Goodman SH. Emory University Project on children of disturbed parents. Schizophr Stata Corp; 1999.
Bull. 1987;13:411-423. 63. Silva PA, Williams S, McGee R. A longitudinal study of children with develop-
29. Yoshida K, Marks MN, Craggs M, Smith B, Kumar R. Sensorimotor and cogni- mental language delay at age 3: later intelligence, reading and behaviour prob-
tive development of infants of mothers with schizophrenia. Br J Psychiatry. 1999; lems. Dev Med Child Neurol. 1987;29:630-640.
175:380-387. 64. Howlin P, Mawhood L, Rutter M. Autism and developmental receptive language
30. Rieder RO, Broman SH, Rosenthal D. The offspring of schizophrenics II, peri- disorder—a follow-up comparison in early adult life II: social, behavioural and
natal factors and IQ. Arch Gen Psychiatry. 1977;34:789-799. psychiatric outcomes. J Child Psychol Psychiatry. 2000;41:561-578.
31. Pasamanick B, Rodgers ME, Lilienfield AM. Pregnancy experience and behav- 65. Dalman C, Allebeck P, Culberg J, Grunewald C, Köster M. Obstetric complica-
iour disorder in children. Am J Psychiatry. 1956;112:613-618. tions and the risk of schizophrenia: a longitudinal study of a national birth co-
32. Sameroff A, Chandler M. Reproductive risk and the continuum of caretaking ca- hort. Arch Gen Psychiatry. 1999;56:234-240.
sualty. In: Horowitz F, ed. Review of Child Development Research. Vol 4. Chi- 66. Jones PB, Rantakallio P, Hartikainen A-L, Isohanni M, Sipila P. Schizophrenia as
cago, Ill: University of Chicago Press; 1975:187-244. a long-term outcome of pregnancy, delivery and perinatal complications: a 28
33. Cicchetti D, Cannon TD. Neurodevelopmental processes in the ontogenesis and year follow up of the 1966 North Finland General Population birth cohort. Am
epigenesis of psychopathology. Dev Psychopathol. 1999;11:375-393. J Psychiatry. 1997;155:355-364.
34. Poulton R, Caspi A, Moffitt TE, Cannon M, Murray RM, Harrington H-L. Chil- 67. Cannon TD, Rosso IM, Hollister JM, Bearden CE, Sanchez LE, Hadley TA. Pro-
dren’s self-reported psychotic symptoms and adult schizophreniform disorder: spective cohort study of genetic and perinatal influences in the etiology of schizo-
a 15-year longitudinal study. Arch Gen Psychiatry. 2000;57:1053-1058. phrenia. Schizophr Bull. 2000;26:249-256.
35. Silva PA, Stanton WR. From Child to Adult: The Dunedin Multidisciplinary Health 68. Rosso IM, Cannon TD, HuttunenT, Huttunen MO, Lönnqvist J, Gasperoni TL. Ob-
and Development Study. Auckland, New Zealand: Oxford University Press; 1996. stetric risk factors for early-onset schizophrenia in a Finnish birth cohort. Am
36. Wright BRE, Caspi A, Moffitt TE, Miech R, Silva PA. Reconsidering the relation- J Psychiatry. 2000;157:801-807.
ship between SES and delinquency: causation but not correlation. Criminology. 69. Zornberg GL, Buka SL, Tsuang MT. Hypoxic-ischaemia-related fetal/neonatal com-
1999;37:175-194. plications and risk of schizophrenia and other nonaffective psychoses: a 19 year
37. Elley, W.B., & Irving, J.C. Revised socio-economic index for New Zealand. N Z J longitudinal study. Am J Psychiatry. 2000;157:196-202.
Educ Studies. 1976;7:153-167. 70. Myhrman A, Rantakallio P, Isohanni M, Jones PB, Partanen U. Does unwanted-
38. Stanton WR, McGee R, Silva PA. Indices of perinatal complications, family back- ness of a pregnancy predict schizophrenia? Br J Psychiatry. 1996;169:637-640.
ground, child rearing and health as predictors of early cognitive and motor de- 71. Jones PB, Murray RM. The genetics of schizophrenia is the genetics of neurode-
velopment. Pediatrics. 1991;88:954-959. velopment. Br J Psychiatry. 1991;158:615-623.
39. Buckfield, P.M. Perinatal events in the Dunedin city population, 1967-1973. N Z 72. Gottesman II. Schizophrenia Genesis: The Origin of Madness. New York, NY: Free-
Med J. 1978;88:244-246. man; 1991.
40. Henry B, Moffitt T, Robins L, Earls F, Silva P. Early family predictors of child and 73. Offord DR, Cross LA. Behavioral antecedents of adult schizophrenia. Arch Gen
adolescent antisocial behaviour: who are the mothers of delinquents? Crim Be- Psychiatry. 1969;21:267-283.
hav Mental Health. 1993;3:97-118. 74. Rutter M. Psychopathology and development, 1: childhood antecedents of adult
41. Touwen BC, Prechtl HR. The neurological examination of the child with minor psychiatric disorder. Aust N Z J Psychiatry. 1984;18:225-234.
nervous dysfunction. In: Clinics in Developmental Medicine. No. 38. London, En- 75. Cannon M, Jones P, Gilvarry C, Rifkin L, McKenzie K, Foerster A, Murray RM.
gland: Heineman; 1970:1-105. Premorbid social functioning in schizophrenia and bipolar disorder: similarities
42. Bayley N. The Bayley Scale of Infant Development. New York, NY: Psychological and differences. Am J Psychiatry. 1997;154:1544-1550.
Corp; 1969. 76. Malmberg A, Lewis G, David A, Allebeck P. Premorbid adjustment and person-
43. McCarthy D. McCarthy Scales of Children’s Abilities. New York, NY: Psychologi- ality in people with schizophrenia. Br J Psychiatry. 1998;172:308-313.
cal Corp; 1972. 77. Davidson M, Reichenberg A, Rabinowitz J, Weiser M, Kaplan Z, Mark M. Behav-
44. Arnheim DD, Sinclair WA. The Clumsy Child. St Louis, Mo: VC Mosby Co; 1974. ioural and intellectual markers for schizophrenia in apparently healthy male ado-
45. Reynell J. Reynell Developmental Language Scales. London, England: National lescents. Am J Psychiatry. 1999;156:1328-1335.
Foundation for Educational Research; 1969. 78. Claridge G. Single indicator of risk for schizophrenia: probable fact or likely myth?
46. Kirk SA, McCarthy JJ, Kirk WD. The Illinois Test of Psycholonguistic Abilities. Schizophr Bull. 1994;20:157-168.
Rev ed. Urbana: University of Illinois Press; 1968. 79. Van Os J, Gilvarry C, Bale R, Van Horn E, Tattan T, White I, Murray RM. A com-
47. Dunn L. The Peabody Picture Vocabulary Test. Minneapolis, Minn: American Guid- parison of the utility of dimensional and categorical representations of psycho-
ance Service; 1995. sis. Psychol Med. 1999;29:595-606.
48. Terman LM, Merrill MR. Stanford Binet Intelligence Scale. Boston, Mass: Hough- 80. Meehl PE. Schizotaxia, schizotypy, schizophrenia. Am Psychol. 1962;17:827-
ton Mifflin; 1960. 838.
49. Weschler D. The Weschler Intelligence Scale for Children–Revised. New York, 81. Meehl PE. Schizotaxia revisited. Arch Gen Psychiatry. 1989;46:935-944.
NY: Psychological Corp; 1974. 82. Tsuang MT, Stone WS, Faraone SV. Towards the prevention of schizophrenia.
Biol Psychiatry. 2000;48:349-356.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 59, MAY 2002 WWW.ARCHGENPSYCHIATRY.COM

456

©2002 American Medical Association. All rights reserved.