Available via license: CC BY 4.0
Content may be subject to copyright.
Psychological Medicine
cambridge.org/psm
Original Article
*Joint senior authors.
Cite this article: Carey E, Healy C, Perry Y,
Gillan D, Whitehouse AJO, Cannon M, Lin A
(2023). Evidence that infant and early
childhood developmental impairments are
associated with hallucinatory experiences:
results from a large, population-based cohort
study. Psychological Medicine 53, 2116–2124.
https://doi.org/10.1017/S0033291721003883
Received: 28 April 2021
Revised: 3 September 2021
Accepted: 3 September 2021
First published online: 29 September 2021
Keywords:
Cognition; Early childhood development;
Hallucinatory experiences; Motor
Author for correspondence:
Eleanor Carey,
E-mail: eleanorcarey1@rcsi.com
© The Author(s), 2021. Published by
Cambridge University Press. This is an Open
Access article, distributed under the terms of
the Creative Commons Attribution licence
(https://creativecommons.org/licenses/by/4.0/),
which permits unrestricted re-use, distribution
and reproduction, provided the original article
is properly cited.
Evidence that infant and early childhood
developmental impairments are associated
with hallucinatory experiences: results from a
large, population-based cohort study
Eleanor Carey1,2 , Colm Healy1, Yael Perry3, Diane Gillan4,
Andrew J. O. Whitehouse3, Mary Cannon1,2,5,*and Ashleigh Lin3, *
1
Department of Psychiatry, Royal College of Surgeons in Ireland, Dublin, Ireland;
2
Trinity College Institute of
Neuroscience, Dublin, Ireland;
3
Telethon Kids Institute, The University of Western Australia, Perth, WA, Australia;
4
Department of Psychology, Beaumont Hospital, Dublin, Ireland and
5
Department of Psychiatry, Beaumont
Hospital, Dublin, Ireland
Abstract
Background. Cognitive and motor dysfunction are hallmark features of the psychosis
continuum, and have been detected during late childhood and adolescence in youth who
report psychotic experiences (PE). However, previous investigations have not explored infancy
and early childhood development. It remains unclear whether such deficits emerge much
earlier in life, and whether they are associated with psychotic, specifically hallucinatory,
experiences (HE).
Methods. This study included data from Gen2 participants of The Raine Study (n= 1101), a
population-based longitudinal cohort study in Western Australia. Five areas of childhood
development comprising: communication; fine motor; gross motor; adaptive (problem-
solving); and personal-social skills, were assessed serially at ages 1, 2 and 3 years.
Information on HE, depression and anxiety at ages 10, 14 and 17 years was obtained.
HE were further subdivided into those with transient or recurrent experiences. Mixed
effects logistic regression models and cumulative risk analyses based on multiple domain
delays were performed.
Results. Early poorer development in multiple areas was noted from ages 1, 2 and 3 years
among youth who reported HE. Early developmental delays significantly increased the risk
for later HE. This association was particularly marked in the recurrent HE group, with
over 40% having early developmental delays in multiple domains. There was no significant
association between early childhood development and later anxiety/depression apart from
lower gross motor scores at age 3.
Conclusions. The findings suggest that early pan-developmental deficits are associated with
later HE, with the effect strongest for young people who report recurrent HE throughout
childhood and adolescence.
Introduction
Poor mental health is being reported at increasingly younger ages, and there has been much
focus traditionally on affective disorders and anxiety (Caspi et al., 2020; Merikangas et al.,
2010). However, there is growing recognition of the importance of psychotic phenomena in
youth. Psychotic experiences (PE), the subclinical expressions of psychotic symptoms, occur
in the absence of a psychotic disorder and mainly concern auditory (both verbal and non-
verbal) and, less commonly, visual hallucinatory experiences (HE) (Kelleher, Harley,
Murtagh, & Cannon, 2011). PE are most prevalent during late childhood, with 17% of 9–12
year olds within the general population reporting these phenomena (Kelleher et al., 2012a),
and a further 12% of children and adolescents reporting auditory hallucinations specifically
(Maijer, Begemann, Palmen, Leucht, & Sommer, 2018). A growing body of research suggests
that PE are not only homotypically or inevitably linked to psychotic disorders (Healy et al.,
2019), but also represent markers of vulnerability for a range of psychopathology and poorer
functional outcomes into adulthood (Carey et al., 2021). PE have also been shown to be asso-
ciated with suicidality and substance abuse in young people (Cederlöf et al., 2017; Honings,
Drukker, Groen, & van Os, 2016; Kelleher, Cederlöf, & Lichtenstein, 2014). For the majority
of young people (75–90%), PE are transient in nature (Calkins et al., 2017), but, in those for
whom PE persist, greater risk of adverse outcomes have been shown (Calkins et al., 2017). The
recurrence of PE therefore may be indicative of a more serious underlying psychopathological
process.
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
Early childhood development, which includes physical, social
and cognitive development, strongly influences mental health
and well-being throughout the lifespan and is critical for influen-
cing a child’s developmental trajectory (Irwin, Siddiqi, &
Hertzman, 2007). Early childhood represents a critical time for
the early detection of risks for later psychopathology (Green
et al., 2019). Cognitive variances in early life lead to a very high
risk of not only developmental issues for children, but also
social-emotional and behavioural problems at later ages (Carter,
Briggs-Gowan, & Davis, 2004), and should be signalled for clinical
intervention.
Investigations of early childhood development and its role in
predicting later schizophrenia have been carried out using birth
cohorts and data linkage approaches but less research has been
carried out with HE as the outcome. Motor deficits during devel-
opment likely represent an endophenotype for schizophrenia
(Burton et al., 2016). Clarke et al. (2011), in a study analysing
data from the Finnish population register, found that delayed
milestones in infancy significantly increased the risk of a later
diagnosis of schizophrenia, in a dose–response manner. Cannon
et al. (2002), in analyses from a New Zealand birth cohort
study, found that subtle but significant impairments in neuromo-
tor, language and cognitive development were detectable from age
3 in children who reported psychotic symptoms at age 11 and
were also detectable in individuals later diagnosed with schizo-
phreniform disorder (but not depression, anxiety or mania) at
age 26. However, developmental data were not available before
the age of 3. Hameed et al. (2018) showed that a decline in
early social and communication skills over the first 4 years of
life, and a declining pattern of fine motor skills in males, was
predictive of later PE (at age 12 years) in a UK birth cohort.
Most studies in this area have focused on childhood and
adolescence, with research showing that cognitive and motor dys-
function at this age is also associated with PE (Carey et al., 2019,
2021; Mollon et al., 2016). A lag in cognitive development from
age 8 years has been detected in children who later report PE
(Gur et al., 2014). However, whether differences in developmental
trajectory can be noted from infancy remains unclear. Questions
pertaining to whether there is an association between early child-
hood developmental delays and poorer outcomes also remain.
Using data from Gen1 and Gen2 of the Raine study, our study
addressed the following aims:
(1) To determine if early childhood development in separate or
multiple domains of development in the first 3 years of life
is associated with HE in childhood and adolescence.
(2) To compare these associations with those of early childhood
development and anxiety/depression in childhood and
adolescence.
(3) To investigate whether the presence of early childhood devel-
opmental delays are associated with recurrence of HE in ado-
lescence, which is a marker for poorer outcomes.
Methods
Participants
Ethical approval for the study was granted by the University of
Western Australia Human Research Ethics Committee (HREC).
Data were taken from The Raine Study, a pregnancy cohort
study that began in 1989, in which 2900 pregnant women were
enrolled with an average recruitment age of 18 weeks of gestation,
from the public antenatal clinic at the major obstetric hospital in
Perth, Western Australia, and nearby private practices. The
ongoing study has followed 2446 infants born over a 3-year
period, with numerous follow-up assessments in childhood and
adolescence. Inclusion criteria for mothers’participation in the
study included gestational age between 16 and 20 weeks, profi-
ciency in English and an intention to remain in Western
Australia so to allow for long-term participation. Demographic
and social information of parents/guardians was collected
throughout the study. The rate of attrition in the Raine study is
29% from age 1 year to age 17 years (Straker et al., 2017).
Participants are representative of the general Western Australian
population at the time of recruitment and into infancy (Straker
et al., 2017).
Measures
Early childhood development
The Infant Monitoring Questionnaire (Bricker, Squires,
Kaminski, & Mounts, 1988) (IMQ; now known as the Ages and
Stages Questionnaire) was completed by parents/guardians at
ages 1, 2 and 3 years. The IMQ is a screening tool for monitoring
early childhood development in five broad domains: communica-
tion, gross motor, fine motor, adaptive (problem-solving) and
personal-social. A 35-item version of the IMQ was administered
at ages 1 and 2 years, and a 30-item version was administered
at age 3 years. Upon agreement with the authors, age 1 and age
2 IMQ data were condensed to 30-items for standardisation
within the analyses, due to the inclusion of one item that is 150
developmental quotient in each domain. This item was therefore
removed from each domain at age 1 and age 2 years, and no sig-
nificant differences were noted. Parents were asked to assess their
child’s behaviour by domain, and score ‘yes’if the child does
behave in such way, ‘sometimes’if the child occasionally behaves
in such way or ‘no’if the child does not behave in such way.
Higher scores indicate better development. The IMQ contain
age-specific developmental items. The IMQ has good criterion
validity, sensitivity, specificity and test–retest reliability (Bricker
et al., 1988). The IMQ also has good criterion validity when com-
pared with the Bayley Scales of Infant Development, and the
Stanford–Binet Intelligence Test (r= 0.89) (Bricker & Squires,
1989) and is endorsed for use by the American Academy of
Pediatrics (Schonhaut, Armijo, Schönstedt, Alvarez, & Cordero,
2013).
The IMQ was scored as per the instructions in the manual
(Squires, 1993) and the suggested cut-off scores were used to
define low-functioning infants and normal-functioning infants
in each domain (see online Supplementary Table S3). Scoring
below this cut-off was labelled as ‘Delay’, while scoring above
this cut-off was labelled as ‘Average’. A further measure of low
functioning was calculated based on scores within the lowest
10th percentile for each domain. The data were also dichotomised
and labelled as ‘Low’(⩽10 percentile) and ‘Average’(⩾11
percentile).
Childhood/adolescent outcomes: hallucinatory experiences and
anxiety/depression
The Child Behaviour Checklist (CBCL)/Youth Self-Report (YSR)
(Achenbach, 1991; Achenbach & Edelbrock, 1983) are question-
naires used to measure emotional and behavioural problems in
children and adolescents, and contain comparable items
(Achenbach, Dumenci, & Rescorla, 2002). They include eight
Psychological Medicine 2117
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
scales: anxious/depressed, depressed, somatic complaints, social
problems, thought problems, attention problems, rule-breaking
behaviour and aggressive behaviour. For the purpose of the cur-
rent analyses, we use CBCL/YSR data collected at ages 10 years
(parent-report) and 14 and 17 (self-report) years.
HE were indexed by two items on the ‘thought problems’sub-
scale of the CBCL/YSR, ‘I hear sounds or voices that other people
think aren’t there’(item 40) and ‘I see things that other people
think aren’t there’(item 70). These items have been found to
have good predictive power for measuring PE (Kelleher et al.,
2011). HE were considered as the endorsement of ‘somewhat/
sometimes true’or ‘very true/often true’for either or both
items, at one or more of the three time-points. Subgroups of
those with transient experiences (reported at one time-point
only) and recurrent experiences (reported at more than one
time-point) were also created.
Anxious/depressed symptomology was determined using
standardised t-scores from the CBCL/YSR anxious/depressed
subscale. A cut-off point for t-scores over and including 70 was
applied, deemed to be clinically significant by Achenbach
(1991). Participants with a t-score of ⩾70 at one or more of the
three time-points (ages 10, 14 and 17 years) and who had never
reported PE were assigned to the anxious/depressed group for
comparison.
Covariates
Covariates entered in the analyses included child’s age (in
months) at the time of their developmental assessment; maternal
highest qualification, birthweight, sex at birth and family income
at the time of the developmental assessments. Maternal highest
qualification was measured when the participants were aged 8
years and was scaled from 0 to 8, and ranged from no qualifica-
tion to postgraduate diploma/higher degree. Family income was
measured at ages 1, 2 and 3 years and was based on five categories
(less than $7000, $7000–11 999, $12 000–23 999, $24 000–35 999
and more than $36 000).
Statistical analysis
Participants were divided into the HE group (those who reported
PE at any time-point) and the anxious/depressed group (those
who scored above the cut-off on the CBCL/YSR at any time-
point). Controls were defined as those who did not report HE
nor had anxious/depressed symptoms above the cut-off at any
time-point. The HE group was subdivided into those with transi-
ent HE (those who reported HE at one time-point only) and
recurrent HE (those who reported PE at either two or three
time-points).
The HE and anxious/depressed groups were compared to con-
trol participants on demographic characteristics using independ-
ent ttests and a chi-squared test. Raw IMQ scores were
converted to z-scores, in order to standardise each version admi-
nistered at ages 1, 2 and 3 years. The data were analysed using
complete case approaches.
Using Stata ver. 16 (StataCorp, 2009), we fitted a series of
mixed-effect logistic models with a random effect accounting
for the within-subject variance. We report the main effects of
early childhood development in the first 3 years of life and later
outcomes in childhood and/or adolescence. The exposure vari-
ables were the five subscales of the IMQ (communication, gross
motor, fine motor, adaptive and personal-social) at ages 1, 2
and 3 years and the primary outcome variables were HE and
anxious/depressed group membership. This approach allowed
for the identification of associations between developmental
scores and HE or anxious/depressed symptoms by detecting
unique variances within the scores after accounting for within-
and between-subject variance. Each of the aforementioned covari-
ates was adjusted for throughout all of the analyses.
Developmental time-specific investigations were then con-
ducted using separate logistic regression models (at ages 1, 2
and 3 years, respectively) to investigate the association between
early childhood development at specific ages and later HE or anx-
ious/depressed symptoms.
Cumulative risk based on multiple domain delays was also
calculated. Cut-off scores for the IMQ were applied at each year
(ages 1, 2 and 3 years). Three categories were created, 0 domain
delays (reference group), 1 or 2 domain delays and 3–5 (multiple)
domain delays, where the outcome variables were HE and anx-
ious/depressed. Cumulative risk at age 3 years based on scores
within the lowest 10th percentile on each domain was also
analysed.
All analyses were also performed to determine associations
between the early childhood developmental scores and subgroups
of HE, transient and recurrent.
Results
Of the 2446 participants who initially took part, data from
n= 1101 participants were available for the present analysis. An
attrition analysis is presented in online Supplementary Table S1.
Those who were included in the present analysis were younger,
had a higher birthweight and better overall development at age
2 years, as well as having mothers with higher qualifications,
than those who were not included.
Hallucinatory experiences
The prevalence of HE was 11.1% in the overall cohort. Among the
participants in the HE group (n= 228), 48% reported at age 17;
38% reported at age 14 and 14% at age 10 (parent-report). At
age 10 years, auditory HE were reported at higher rates (83%)
while at 14 years, visual HE were more prevalent (84%). At age
17 years, rates of auditory and visual HE were comparable (50%
respectively), and 48% of those who reported HE at this age
reported both. Transient experiences were reported in 82% of
cases, while recurrent experiences were reported in 18% of
cases. A total of n= 40 participants reported HE at two time-
points, and a further n= 2 participants reported HE at three
time-points.
Anxious/depressed
The prevalence of anxious/depressed symptoms at or above cut-
off was 5.6% in the overall cohort. A total of n=114 participants
scored on or above cut-off at least once at ages 10, 14 or 17 years.
Seven (6%) had symptoms at more than one time-point. A total of
32% had symptoms at age 17; 18% had symptoms at age 14 and
32% had symptoms at age 10. Of note, n=43 (37.7%) of the anx-
ious/depressed group also reported PE, and were therefore only
included in the HE group.
No significant group differences in demographic characteris-
tics were found between the HE and control groups, with
the exception of maternal highest level of education (t=−3.80,
p< 0.01) and family income at age 1 (t=−3.76, p< 0.01), where
2118 Eleanor Carey et al.
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
less maternal education and a lower family income at age 1 were
noted for the HE group. No significant group differences were
found between the anxious/depressed and control groups (see
Table 1).
Association between early childhood development and PE
Mean scores for the IMQ subscales at each age (1, 2 and 3 years)
for the HE, anxious/depressed and control groups, including the
subgroups of the HE group (transient and recurrent HE), are pre-
sented in online Supplementary Table S2.
In the overall random effects model, which accounted for
development over the first 3 years of age, there was no significant
association between any of developmental domains and HE. In
the age-specific analysis, lower adaptive scores at age 1 year
were associated with HE. At age 2 years, lower scores in 4 of
the 5 developmental domains (with the exception of gross
motor) were associated with HE. At age 3 years, lower scores in
any of the five domains (communication, gross motor, fine
motor, adaptive and personal-social) were associated with HE
(see Table 2).
These results are presented, stratified by gender, in Tables S4
and S5 of the online Supplementary materials.
Association between early childhood development and
anxious/depressed symptoms
No significant associations between any of developmental
domains and anxious/depressed symptoms were found. In the
age-specific analysis, only lower gross motor scores at age 3
years were associated with anxious/depressed symptoms
(see Table 2).
Association between early childhood development and
transient and recurrent HE
No significant associations between any of developmental
domains and transient HE were found. In the age-specific ana-
lysis, lower adaptive scores at ages 1, 2 and 3 years, as well as
lower fine motor scores at age 3 years, were significantly asso-
ciated with transient HE.
Table 1. Demographic characteristics for the HE group and anxious/depressed groups when compared to the control group
Descriptive statistics Analysis (reference: control group)
HE group
Mean (S.D.)
Anxious/depressed
group
Mean (S.D.)
Control group
Mean (S.D.)
HE group
χ
2
or t p-value
Anxious/depressed
group
χ
2
or t p-value
Demographic variable
Sex at birth, n(%) 1.03 0.31 1.66 0.20
Males 119 (52%) 40 (56%) 388 (48%)
Females 109 (49%) 31 (44%) 414 (52%)
Ethnicity
a
Caucasian (European descent) 203 (91%) 69 (99%) 703 (90%)
Aboriginal 1 0 5
Polynesian 0 0 8
Vietnamese 0 0 4
Chinese 13 0 37
Indian 6 1 20
Other 0 0 7
Birthweight (in g) 3281.7 (671.8) 3281.7 (626.2) 3341.3 (560.1) −1.35 0.18 −0.09 0.93
Age 1 (months) 1.50 (0.11) 1.14 (0.11) 1.14 (0.10) −0.70 0.48 0.21 0.84
Age 2 2.13 (0.15) 2.16 (0.18) 2.13 (0.14) 0.07 0.94 −1.46 0.14
Age 3 3.15 (0.16) 3.16 (0.18) 3.12 (0.12) −3.13 0.002 −2.94 0.003
Age 10 10.59 (0.20) 10.62 (0.17) 10.57 (0.17) −1.39 0.17 −2.14 0.03
Age 14 13.99 (0.21) 14.02 (0.23) 13.99 (0.19) −0.70 0.48 −1.26 0.21
Age 17 17.07 (0.27) 17.03 (0.29) 17.02 (0.22) −2.80 0.005 −0.39 0.70
Maternal highest qualification 2.42 (2.49) 2.60 (2.35) 3.14 (2.30) −3.80 <0.001 −1.84 0.07
Family income age 1 2.77 (1.20) 2.78 (1.40) 3.11 (1.24) −3.76 <0.001 −1.90 0.06
Family income age 2 2.83 (1.70) 3.03 (2.11) 3.05 (1.41) −1.61 0.11 −0.07 0.94
Family income age 3 2.95 (1.63) 2.80 (1.61) 3.10 (1.32) −1.23 0.22 −1.74 0.08
S.D., standard deviation.
a
Some missing data.
Psychological Medicine 2119
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
Lower communication scores in the first 3 years were asso-
ciated with recurrent HE. In the age-specific analysis, lower
gross motor and fine motor scores at age 1 year were significantly
associated with recurrent HE. At age 2 years, lower scores in 4 of
the 5 developmental domains (with the exception of gross motor)
were associated with recurrent HE. At age 3 years, lower scores in
4 of the 5 early childhood development domains (communica-
tion, gross motor, adaptive and personal-social) were significantly
associated with recurrent HE. All data are presented in Fig. 1.
Cumulative risk based on multiple domain delays
Based on cut-off scores, having 3–5 developmental domain delays
at age 2 led to an almost 5-fold increased risk for HE (adjusted
odds ratio 4.86, 95% confidence interval 1.69–14.05, p= 0.004).
Having 1–2 developmental domain delays at age 3 led to an
almost 2-fold increased risk for HE and having multiple domain
delays increased the risk of HE over 3-fold (see Table 3). Among
young people who reported recurrent HE, 30.9% had develop-
mental domain delays at age 2 years and 40.5% had developmen-
tal domain delays at age 3 years.
In the percentile analysis, delays (lowest 10th percenile for all
participants) in 1–2 domains led to over a 2-fold increase risk for
HE, while delays in multiple domains led to over a 3-fold
increased risk for HE (see Table 4). For the HE subgroups, delays
(by percentile) in 1–2 developmental domains led to over a 2-fold
increased risk for transient HE. Delays in 3–5 developmental
domains led to over a 7-fold increased risk for recurrent HE.
Multiple developmental domain delays did not lead to a
significant increase in cumulative risk for anxious/depressed
symptoms, in either the cut-off or percentile analysis.
Discussion
In this study, we examined longitudinally the association between
early childhood development and HE in childhood and adoles-
cence in a population-based, representative sample of young peo-
ple. Participants who reported HE in childhood and/or
adolescence showed lower early developmental scores from ages
1 and 2 years on a range of measures. By age 3 years, lower scores
in any of the five developmental domains were significantly asso-
ciated with the risk of later reporting HE. Cumulative risk based
on multiple (3–5) domain delays at age 3 years led to over a 3-fold
increased risk for HE. For those participants who would later
report clinically significant anxious/depressed symptomology,
no association was found for overall development over the first
3 years, with the exception of gross motor scores at age 3, and
there was also no increased cumulative risk based on multiple
domains delays.
Further analyses within the HE group showed that when sub-
dividing the group by transient and recurrent experiences, lower
adaptive scores at ages 1, 2 and 3 years were specifically associated
Table 2. Overall and age-specific developmental scores between ages 1 and 3 and risk of later HE (n= 228) [also subdivided into transient (n= 114) or recurrent
(n= 42)] and anxiety/depression (n= 71) (compared with Controls (n= 802) (reference group is controls)
Communication Gross motor Fine motor Adaptive Personal-social
OR (CI) OR (CI) OR (CI) OR (CI) OR (CI)
HE group
Mixed effect model 1.04 (0.92–1.19) 1.02 (0.90–1.07) 0.99 (0.84–1.16) 1.05 (0.92–1.05) 1.01 (0.87–1.18)
Age 1 year 1.14 (0.96–1.35) 1.09 (0.93–1.27) 1.08 (0.93–1.28) 1.27 (1.08–1.49) 1.10 (0.94–1.30)
Age 2 years 1.39 (1.18–1.64) 1.16 (0.99–1.39) 1.23 (1.03–1.47) 1.39 (1.18–1.66) 1.20 (1.01–1.45)
Age 3 years 1.32 (1.12–1.54) 1.25 (1.06–1.47) 1.32 (1.14–1.54) 1.35 (1.16–1.56) 1.20 (1.02–1.43)
Anxious/depressed group
Mixed effect model 0.98 (0.63–1.54) 1.19 (0.78–1.79) 1.12 (0.72–1.75) 1.17 (0.76–1.82) 1.19 (0.76–1.85)
Age 1 year 1.02 (0.77–1.35) 0.91 (0.68–1.22) 1.01 (0.76–1.33) 1.01 (0.76–1.35) 1.20 (0.93–1.56)
Age 2 years 0.92 (0.65–1.28) 1.20 (0.93–1.54) 1.20 (0.92–1.58) 1.30 (1.00–1.72) 1.08 (0.79–1.47)
Age 3 years 1.05 (0.79–1.39) 1.39 (1.11–1.75) 1.20 (0.93–1.54) 1.14 (0.88–1.47) 1.19 (0.90–1.56)
Transient HE group
Mixed effect model 1.03 (0.60–1.78) 1.05 (0.63–1.75) 1.02 (0.62–1.66) 1.28 (0.85–1.92) 1.08 (0.72–1.59)
Age 1 year 1.05 (0.84–1.32) 0.94 (0.76–1.19) 0.89 (0.70–1.12) 1.33 (1.08–1.64) 1.00 (0.80–1.23)
Age 2 years 1.12 (0.88–1.43) 1.06 (0.85–1.33) 0.97 (0.76–1.25) 1.28 (1.02–1.61) 1.05 (0.82–1.35)
Age 3 years 1.14 (0.93–1.41) 1.16 (0.93–1.45) 1.33 (1.00–1.61) 1.27 (1.04–1.54) 1.15 (0.92–1.43)
Recurrent HE group
Mixed effect model 2.04 (1.14–3.70) 1.72 (0.79–3.85) 1.72 (0.96–3.03) 1.37 (0.78–2.43) 1.47 (0.81–2.70)
Age 1 year 1.39 (0.99–1.92) 1.37 (1.06–1.75) 1.43 (1.09–1.89) 1.19 (0.86–1.63) 1.30 (0.95–1.78)
Age 2 years 2.17 (1.54–3.03) 1.28 (0.90–1.82) 2.04 (1.43–2.94) 1.44 (1.01–2.08) 1.56 (1.06–2.27)
Age 3 years 1.85 (1.45–2.38) 1.64 (1.27–2.17) 1.82 (0.98–1.82) 1.41 (1.06–1.92) 1.59 (1.14–2.17)
OR, odds ratio; CI, confidence intervals.
Bolded text represents analyses where 95% CI do not cross 1.
Odds ratios are adjusted for sex at birth, age, birthweight, maternal highest qualification and family income at each year.
2120 Eleanor Carey et al.
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
with transient HE. Delays in 1 or 2 developmental domains at age
3 years led to over a 2-fold increased risk for transient HE.
Stronger effects were found for those who reported recurrent
HE, in that a greater number of associations were found with
early developmental scores compared to those who only reported
HE at one time-point in childhood or adolescence. Both at ages 2
and 3 years, lower scores in 4 of the 5 developmental domains
were associated with recurrent HE. At age 3 years, over 40% of
the recurrent HE group had at least one developmental delay, as
measured by cut-off scores. Multiple (3–5) domain delays led to
over a 7-fold increased risk. These findings convey the strong
associations between poorer early childhood development and
mental health outcomes later in life. Our findings indicate a spe-
cific risk associated with the psychosis spectrum, and a strong
association with recurrent HE throughout childhood and adoles-
cent development.
Fig. 1. Longitudinal unadjusted mean z-scores with standard deviation bars for each developmental domain (ages 1, 2 and 3 years) for all groups.
Table 3. Cumulative risk for later HE (n= 228) [Transient (n= 114) or recurrent (n= 42)] and anxiety/depression (n= 71) based on number of developmental delays
(below cut-off) at ages 1, 2 and 3 years for each group
Cumulative risk
(0 delays reference category)
1–2 delays
N(% below cut-off) OR (CI) p-value
3–5 delays
N(% below cut-off) OR (CI) p-value
Age 1
HE group 41 (18) 1.70 (1.10–2.62) 0.02 4 (1.8) 0.75 (0.16–3.52) 0.72
Anxious/depressed group 8 (11.3) 0.85 (0.37–1.95) 0.70 0 ––
Transient HE group
a
17 (14.9) 1.46 (0.82–2.61) 0.20 1 (0.9) 0.75 (0.09–5.99) 0.78
Recurrent HE group
a
9 (21.4) 1.88 (0.83–4.23) 0.13 0 ––
Controls 86 (10.7) 10 (1.2)
Age 2
HE group 50 (21.9) 1.81 (1.20–2.72) 0.004 8 (3.5) 4.86 (1.69–14.05) 0.004
Anxious/depressed group 10 (14.1) 1.00 (0.49–2.08) 0.99 1 (1.4) 1.66 (0.19–14.08) 0.64
Transient HE group
a
20 (17.5) 1.24 (0.72–2.13) 0.45 1 (0.9) 0.95 (0.11–8.02) 0.96
Recurrent HE group
a
10 (23.8) 3.49 (1.46–8.33) 0.005 3 (7.1) 21.11 (4.22–105.91) <0.001
Controls 118 (14.7) 7 (0.9)
Age 3
HE group 60 (26.3) 1.80 (1.24–2.62) 0.002 17 (7.5) 3.60 (1.70–7.23) 0.001
Anxious/depressed group 18 (25.4) 1.53 (0.84–2.78) 0.17 3 (4.2) 1.31 (0.29–5.97) 0.72
Transient HE group
a
38 (33.3) 2.17 (1.38–3.42) 0.001 3 (2.6) 1.33 (0.37–4.70) 0.66
Recurrent HE group
a
10 (23.8) 1.45 (0.64–3.31) 0.37 7 (16.7) 7.68 (2.73–21.58) <0.001
Controls 154 (19.2) 19 (2.4)
a
Subgroups of the HE group.
Psychological Medicine 2121
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
The present study has extended previous research on early
childhood development and PE. Using the Denver Developmental
Screening Test (which includes four measures of early develop-
ment: fine and gross motor, social and communication skills),
Hameed et al. (2018) showed that declining communication
and social skills in the first 4 years of life were significantly asso-
ciated with PE at one time-point at age 12 years. Our study has
extended these findings by measuring adaptive (problem-solving)
skills, and by examining HE over three time-points in childhood
and adolescence. In addition, we explored the comparison
between the associations between poorer early development and
later HE and clinically significant anxious/depressed symptomol-
ogy as outcomes. Lower scores in gross motor development at age
3 years were associated with later clinically significant anxious/
depressed symptomology. Although the co-occurrence of PE
and anxiety/depression has been established (Wigman et al.,
2011), our findings suggest that poorer development is driven
by variability associated with psychosis spectrum symptomology
rather than psychopathology more generally, a pattern that has
also been shown in other studies (Cannon et al., 2002).
Our findings pertaining to early infant communication devel-
opment and its relationship with HE, in particular recurrent HE,
are in-line with those of Hameed et al. (2018). Impairments in
language abilities and the dysconnectivity of language networks
have been a central topic in the research of cognitive dysfunction
in psychosis, and are thought to contribute to the development of
auditory hallucinations (Benetti et al., 2015). Altered white matter
in language areas has also been seen in young people with PE
(Dooley et al., 2020). Our results suggest that language and com-
munication deficits may be detectable as early as the first year of
life in young people at risk for psychosis.
Poorer early adaptive development or problem-solving skills
were also associated with HE, and in particular, transient HE.
The transient HE group were characterised by persistently lower
scores in adaptive skills throughout the first 3 years. Reasoning,
control and planning are core to adaptive development and delays
in these areas have been described in the later stages of childhood
in individuals who go on to develop schizophrenia (Reichenberg
et al., 2009). Transient HE, or once-off PE, are linked to poorer
global functioning throughout adolescence and early adulthood
(Healy et al., 2018).
The recurrent HE group had poorer developmental scores in
both cognitive and motor domains throughout the first 3 years
of life. Motor dysfunction may be reflective of a core feature of
psychosis and schizophrenia. Psychomotor developmental devia-
tions are more marked in those later diagnosed with schizophre-
nia, when compared those later diagnosed with bipolar disorder,
and have an earlier onset (Parellada, Gomez-Vallejo, Burdeus, &
Arango, 2017). Keskinen et al. (2015) hypothesised that delayed
motor development may be a marker of other risk processes
that interact with genetic liability for psychosis and schizophrenia.
Murray et al. (2006) retrospectively found that individuals with
schizophrenia had early delays in neuromotor milestones.
Motor development therefore may be an important trait marker
for recurrent HE, which is of important clinical utility, as young
people with persistent symptoms have poorer functional out-
comes (Healy et al., 2018), higher levels of stress reactivity
(Collip et al., 2013) and are at greater risk of conversion to psych-
osis (Dominguez, Wichers, Lieb, Wittchen, & van Os, 2011).
Recurrence is thought to be indicative of a more serious under-
lying psychopathological process, and represents an important,
early, non-specific marker of later mental health problems
(Kelleher et al., 2012b).
Strengths of the study include the availability of very early
developmental data from infancy from a population-based longi-
tudinal birth cohort and the use of a robust and widely-used
assessment of early childhood development, as well as the con-
sistency of measures used across time-points. The HE group
also included those with clinically significant anxious/depressed
symptomology, due to the established co-occurrence (Kounali
et al., 2014) and known overlap with psychopathology more gen-
erally (Trotta et al., 2020). Kounali and colleagues also found that
markers of abnormal neurodevelopment showed stronger associ-
ation with PE, when compared to depression. However, limita-
tions of the study should also be noted. The measurement of
HE was confined to the self-report of two types of symptomology
(namely, auditory and visual HE); however, these are the most
common and validated in childhood community samples
(Laurens, Hobbs, Sunderland, Green, & Mould, 2012). HE were
indexed by responses ‘somewhat/sometimes true’or ‘very true/
often true’which may have been inflated prevalence rates.
Furthermore, at age 10 years, a parent-report of HE was used
which may have led to an underestimation of HE. The prevalence
of clinically significant anxious/depressed symptoms might also
be underestimated, particularly as a high proportion who also
reported HE were counted within the HE group rather than the
Anxious/depressed group as this was the main focus of our ana-
lysis. Within the analyses, the omission of parental psychopath-
ology as a covariate must also be noted. This was not
consistently measured throughout The Raine Study.
Furthermore, the impact of multiple comparisons should be con-
sidered, particularly for the results in the anxious/depressed
group. Finally, attrition rates within The Raine Study may have
led to follow-up bias; however, this may have led to underestima-
tion rather than over estimation of effect sizes since those who
were lost to follow-up had poorer development at age 2 years.
Table 4. Cumulative risk for later HE (transient or recurrent) and anxious/depressed based on number of domain delays at age 3 years for each group
Cumulative risk (0 delays reference category)
1–2 domains
OR (CI) p-value
3–5 domains
OR (CI) p-value
HE group 2.21 (1.54–3.18) <0.001 3.60 (1.82–6.99) <0.001
Anxious/depressed group 1.57 (0.87–2.81) 0.13 1.07 (0.24–4.78) 0.93
Transient HE group
a
2.31 (1.48–3.60) <0.001 1.54 (0.51–4.67) 0.16
Recurrent HE group
a
1.98 (0.92–4.25) 0.08 7.23 (2.62–19.98) <0.001
OR, odds ratio; CI, confidence intervals.
Statistically significant results ( p< 0.05) are presented in bold text.
a
Subgroups of the HE group.
2122 Eleanor Carey et al.
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
In conclusion, our study has shown that cognitive and motor
developmental delays can be detected from infancy in young
people who go on to report HE in childhood and adolescence.
Our findings are reminiscent of the ‘early insult’theory in psych-
osis, where early genetic or environmental insults lead to aberrant
brain development (Murray & Lewis, 1987) which is exacerbated
thereafter by childhood trauma and adversity (Croft et al., 2019).
Exploration of early childhood development as both a meaningful
marker for later mental health outcomes and a viable critical per-
iod for early intervention is an exciting emerging area in psychosis
research.
Supplementary material. The supplementary material for this article can
be found at https://doi.org/10.1017/S0033291721003883.
Acknowledgements. We are grateful to the Raine study participants and their
families and we thank the Raine study staff for cohort coordination and data col-
lection. The core management of the Raine study is funded by The University of
Western Australia, Curtin University, Telethon Kids Institute, Women and
Infants Research Foundation, Edith Cowan University, Murdoch University,
The University of Notre Dame Australia and the Raine Medical Research
Foundation. The study is funded by the National Health and Medical
Research Council (NHMRC).
Author contributions. Each author contributed significantly to the
manuscript. E. Carey (corresponding author) had full access to the data and
had complete freedom to direct its analysis and reporting. C. Healy, A. Lin
and M. Cannon were involved in the conception and design of the study, as
well as the analysis of the data and interpretation of the results. Y. Perry,
D. Gillan and A. Whitehouse contributed to the preparing and editing of previ-
ous drafts of the manuscript and the critical revision of important intellectual
content. All authors gave their approval of the present version to be published.
Financial support. EC, CH and MC are funded by a European Research
Council Consolidator Award to MC (724809 iHEAR). AL is funded by a
National Health and Medical Research Council Career Development
Fellowship (no. 1148793). YP is funded by a fellowship from the Giorgetta
Charity Fund. AW is supported by an Investigator Grant from the National
Health and Medical Research Council (no. 1173896).
Conflict of interest. None.
Role of the funder or sponsor. The funders had no role in the design and
conduct of the study; collection, management, analysis and interpretation of
the data. The Raine study approved final manuscript and agreed upon its sub-
mission for publication.
References
Achenbach, T.M. (1991). Manual for the youth self-report and 1991 profile (pp.
1–13). Burlington: Department of Psychiatry, University of Vermont.
Achenbach, T. M., Dumenci, L., & Rescorla, L. A. (2002). Ten-year comparisons
of problems and competencies for national samples of youth: Self, parent, and
teacher reports. Journal of Emotional and Behavioral Disorders,10(4), 194–203.
Achenbach, T. M., & Edelbrock, C. S. (1983). Manual for the child behavior
checklist and revised child behavior profile.
Benetti, S., Pettersson-Yeo, W., Allen, P., Catani, M., Williams, S., Barsaglini,
A., …Mechelli, A. (2015). Auditory verbal hallucinations and brain dyscon-
nectivity in the perisylvian language network: A multimodal investigation.
Schizophrenia Bulletin,41(1), 192–200.
Bricker, D., & Squires, J. (1989). The effectiveness of parental screening of
at-risk infants: The infant monitoring questionnaires. Topics in Early
Childhood Special Education,9(3), 67–85.
Bricker, D, Squires, J, Kaminski, R, & Mounts, L. (1988). The validity, reliabil-
ity, and cost of a parent-completed questionnaire system to evaluate at-risk
infants. Journal of pediatric psychology,13(1), 55–68.
Burton, B. K., Hjorthøj, C., Jepsen, J. R., Thorup, A., Nordentoft, M., &
Plessen, K. J. (2016). Research review: Do motor deficits during
development represent an endophenotype for schizophrenia? A
meta-analysis. Journal of Child Psychology and Psychiatry,57(4), 446–456.
Calkins, M. E., Moore, T. M., Satterthwaite, T. D., Wolf, D. H., Turetsky, B. I.,
Roalf, D. R., …Gur, R. C. (2017). Persistence of psychosis spectrum
symptoms in the Philadelphia neurodevelopmental cohort: A prospective
two-year follow-up. World Psychiatry,16(1), 62–76.
Cannon, M., Caspi, A., Moffitt, T. E., Harrington, H., Taylor, A., Murray, R.
M., & Poulton, R. (2002). Evidence for early-childhood, pan-developmental
impairment specific to schizophreniform disorder: Results from a longitu-
dinal birth cohort. Archives of General Psychiatry,59(5), 449–456.
Carey, E., Dooley, N., Gillan, D., Healy, C., Coughlan, H., Clarke, M., …Cannon,
M. (2019). Fine motor skill and processing speed deficits in young people
with psychotic experiences: A longitudinal study. Schizophrenia Research,
204, 127–132.
Carey, E., Gillan, D., Healy, C., Dooley, N., Campbell, D., McGrane, J., …
Cannon, M. (2021). Early adult mental health, functional and neuropsycho-
logical outcomes of young people who have reported psychotic experiences:
A 10-year longitudinal study. Psychological Medicine,51(11), 1861–1869.
Carter, A. S., Briggs-Gowan, M. J., & Davis, N. O. (2004). Assessment of young
children’s social-emotional development and psychopathology: Recent
advances and recommendations for practice. Journal of Child Psychology
and Psychiatry,45(1), 109–134.
Caspi, A., Houts, R. M., Ambler, A., Danese, A., Elliott, M. L., Hariri, A., …
Ramrakha, S. (2020). Longitudinal assessment of mental health disorders
and comorbidities across 4 decades among participants in the Dunedin
birth cohort study. JAMA Network Open,3(4), e203221–e203221.
Cederlöf, M., Kuja-Halkola, R., Larsson, H., Sjölander, A., Östberg, P.,
Lundström, S., …Lichtenstein, P. (2017). A longitudinal study of adolescent
psychotic experiences and later development of substance use disorder and
suicidal behavior. Schizophrenia Research,181,13–16.
Clarke, M. C., Tanskanen, A., Huttunen, M., Leon, D. A., Murray, R. M., Jones,
P. B., & Cannon, M. (2011). Increased risk of schizophrenia from additive
interaction between infant motor developmental delay and obstetric com-
plications: Evidence from a population-based longitudinal study.
American Journal of Psychiatry,168(12), 1295–1302.
Collip, D., Wigman, J. T., Myin-Germeys, I., Jacobs, N., Derom, C., Thiery, E.,
…van Os, J. (2013). From epidemiology to daily life: Linking daily life
stress reactivity to persistence of psychotic experiences in a longitudinal
general population study. PLoS ONE,8(4), e62688.
Croft, J., Heron, J., Teufel, C., Cannon, M., Wolke, D., Thompson, A., …
Zammit, S. (2019). Association of trauma type, age of exposure, and fre-
quency in childhood and adolescence with psychotic experiences in early
adulthood. JAMA Psychiatry,76(1), 79–86.
Dominguez, M., Wichers, M., Lieb, R., Wittchen, H.-U., & van Os, J. (2011).
Evidence that onset of clinical psychosis is an outcome of progressively
more persistent subclinical psychotic experiences: An 8-year cohort study.
Schizophrenia Bulletin,37(1), 84–93.
Dooley, N., O’Hanlon, E., Healy, C., Adair, A., McCandless, C., Coppinger, D.,
…Frodl, T. (2020). Psychotic experiences in childhood are associated with
increased structural integrity of the left arcuate fasciculus –A population-
based case-control study. Schizophrenia Research,215, 378–384.
Green, M. J., Tzoumakis, S., Laurens, K. R., Dean, K., Kariuki, M., Harris, F.,
…Carr, V. J. (2019). Early developmental risk for subsequent childhood
mental disorders in an Australian population cohort. Australian & New
Zealand Journal of Psychiatry,53(4), 304–315.
Gur, R. C., Calkins, M. E., Satterthwaite, T. D., Ruparel, K., Bilker, W. B.,
Moore, T. M., …Gur, R. E. (2014). Neurocognitive growth charting in
psychosis spectrum youths. JAMA Psychiatry,71(4), 366–374.
Hameed, M. A., Lingam, R., Zammit, S., Salvi, G., Sullivan, S., & Lewis, A. J.
(2018). Trajectories of early childhood developmental skills and early ado-
lescent psychotic experiences: Findings from the ALSPAC UK birth cohort.
Frontiers in Psychology,8, 2314.
Healy,C.,Brannigan,R.,Dooley,N.,Coughlan,H.,Clarke,M.,Kelleher,I.,…
Cannon, M. (2019). Childhood and adolescent psychotic experiences and
risk of mental disorder: A systematic review and meta-analysis. Psychological
Medicine,49(10), 1589–1599.
Healy, C., Campbell, D., Coughlan, H., Clarke, M., Kelleher, I., & Cannon, M.
(2018). Childhood psychotic experiences are associated with poorer global
Psychological Medicine 2123
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
functioning throughout adolescence and into early adulthood. Acta
Psychiatrica Scandinavica,138(1), 26–34.
Honings, S., Drukker, M., Groen, R., & van Os, J. (2016). Psychotic experiences
and risk of self-injurious behaviour in the general population: A systematic
review and meta-analysis. Psychological Medicine,46(2), 237–251.
Irwin,L.G.,Siddiqi,A.,&Hertzman,G.(2007).Early child development: A power-
ful equalizer. Human Early Learning Partnership (HELP) Vancouver, BC.
Kelleher, I., Cederlöf, M., & Lichtenstein, P. (2014). Psychotic experiences as a
predictor of the natural course of suicidal ideation: A Swedish cohort study.
World Psychiatry,13(2), 184–188.
Kelleher, I., Connor, D., Clarke, M. C., Devlin, N., Harley, M., & Cannon, M.
(2012a). Prevalence of psychotic symptoms in childhood and adolescence:
A systematic review and meta-analysis of population-based studies.
Psychological Medicine,42(9), 1857–1863.
Kelleher, I., Harley, M., Murtagh, A., & Cannon, M. (2011). Are screening
instruments valid for psychotic-like experiences? A validation study of
screening questions for psychotic-like experiences using in-depth clinical
interview. Schizophrenia Bulletin,37(2), 362–369.
Kelleher, I., Keeley, H., Corcoran, P., Lynch, F., Fitzpatrick, C., Devlin, N., …
Harley, M. (2012b). Clinicopathological significance of psychotic experi-
ences in non-psychotic young people: Evidence from four population-based
studies. The British Journal of Psychiatry,201(1), 26–32.
Keskinen, E., Marttila, A., Marttila, R., Jones, P., Murray, G., Moilanen, K., …
Jääskeläinen, E. (2015). Interaction between parental psychosis and early
motor development and the risk of schizophrenia in a general population
birth cohort. European Psychiatry,30(6), 719–727.
Kounali, D., Zammit, S., Wiles, N., Sullivan, S., Cannon, M., Stochl, J., …Heron,
J. (2014). Common versus psychopathology-specific risk factors for psychotic
experiences and depression during adolescence. Psychological Medicine,44
(12), 2557–2566.
Laurens,K.R.,Hobbs,M.,Sunderland,M.,Green,M.J.,&Mould,G.(2012).
Psychotic-like experiences in a community sample of 8000 children aged 9
to 11 years: An item response theory analysis. Psychological Medicine,42(7),
1495–1506.
Maijer, K., Begemann, M. J., Palmen, S. J., Leucht, S., & Sommer, I. E. (2018).
Auditory hallucinations across the lifespan: A systematic review and
meta-analysis. Psychological Medicine,48(6), 879–888.
Merikangas, K. R., He, J.-P., Burstein, M., Swanson, S. A., Avenevoli, S., Cui, L.,
…Swendsen, J. (2010). Lifetime prevalence of mental disorders in US ado-
lescents: Results from the national comorbidity survey replication–adoles-
cent supplement (NCS-A). Journal of the American Academy of Child &
Adolescent Psychiatry,49(10), 980–989.
Mollon,J.,David,A.S.,Morgan,C.,Frissa,S., Glahn,D.,Pilecka,I.,…
Reichenberg, A. (2016). Psychotic experiences and neuropsychological
functioning in a population-based sample. JAMA Psychiatry,73(2), 129–138.
Murray, G., Jones, P., Moilanen, K., Veijola, J., Miettunen, J., Cannon, T., &
Isohanni, M. (2006). Infant motor development and adult cognitive func-
tions in schizophrenia. Schizophrenia Research,81(1), 65–74.
Murray, R. M., & Lewis, S. W. (1987). Is schizophrenia a neurodevelopmental
disorder? British Medical Journal (Clinical Research Ed.),295(6600), 681.
Parellada, M., Gomez-Vallejo, S., Burdeus, M., & Arango, C. (2017).
Developmental differences between schizophrenia and bipolar disorder.
Schizophrenia Bulletin,43(6), 1176–1189.
Reichenberg, A., Caspi, A., Harrington, H., Houts, R., Keefe, R. S., Murray, R.
M., …Moffitt, T. E. (2009). Static and dynamic cognitive deficits in child-
hood preceding adult schizophrenia: A 30-year study. American Journal of
Psychiatry,167(2), 160–169.
Schonhaut, L., Armijo, I., Schönstedt, M., Alvarez, J., & Cordero, M. (2013).
Validity of the ages and stages questionnaires in term and preterm infants.
Pediatrics,131(5), e1468–e1474.
Squires, J. (1993). Infant/child monitoring questionnaires procedures manual.
Oregon: University of Oregon Center of Human Development.
Straker, L., Mountain, J., Jacques, A., White, S., Smith, A., Landau, L., …
Eastwood, P. (2017). Cohort profile: The western Australian pregnancy
cohort (Raine) study–generation 2. International Journal of Epidemiology,
46(5), 1384–1385j.
Trotta, A., Arseneault, L., Caspi, A., Moffitt, T. E., Danese, A., Pariante, C., …
Fisher, H. L. (2020). Mental health and functional outcomes in young adult-
hood of children with psychotic symptoms: A longitudinal cohort study.
Schizophrenia Bulletin,46(2), 261–271.
Wigman, J., Lin, A., Vollebergh, W. A., van Os, J., Raaijmakers, Q. A., Nelson,
B., …Yung, A. (2011). Subclinical psychosis and depression: Co-occurring
phenomena that do not predict each other over time. Schizophrenia
Research,130(1–3), 277–281.
2124 Eleanor Carey et al.
https://doi.org/10.1017/S0033291721003883 Published online by Cambridge University Press
