Neurodevelopmental dimensional assessment of young children at high genomic risk of neuropsychiatric conditions

Abstract Background Individuals with 22q11.2 deletion are at considerably increased risk of neurodevelopmental and psychiatric conditions. There have been very few studies investigating how this risk manifests in early childhood and what factors may underlie developmental variability. Insights into this can elucidate transdiagnostic markers of risk that may underlie later development of neuropsychiatric outcomes. Methods Thirty two children with 22q11.2 Deletion Syndrome (22q11.2DS) (mean age = 4.1 [SD = 1.2] years) and 12 sibling controls (mean age = 4.1 [SD = 1.5] years) underwent in‐depth dimensional phenotyping across several developmental domains selected as being potential early indicators of neurodevelopmental and psychiatric liability. Comparisons were conducted of the dimensional developmental phenotype of 22q11.2DS and sibling controls. For autistic traits, both parents and children were phenotyped using the Social Responsiveness Scale. Results Young children with 22q11.2DS exhibited large impairments (Hedge's g ≥ 0.8) across a range of developmental domains relative to sibling controls, as well as high rates of transdiagnostic neurodevelopmental and psychiatric traits. Cluster analysis revealed a subgroup of children with 22q11.2DS (n = 16; 53%) in whom neurodevelopmental and psychiatric liability was particularly increased and who differed from other children with 22q11.2DS and non‐carrier siblings. Exploratory analyses revealed that early motor and sleep impairments indexed liability for neurodevelopmental and psychiatric outcomes. Maternal autism trait scores were predictive of autism traits in children with 22q11.2DS (intraclass correlation coefficients = 0.47, p = 0.046, n = 31). Conclusions Although psychiatric conditions typically emerge later in adolescence and adulthood in 22q11.2DS, our exploratory study was able to identify a range of early risk indicators. Furthermore, findings indicate the presence of a subgroup who appeared to have increased neurodevelopmental and psychiatric liability. Our findings highlight the scope for future studies of early risk mechanisms and early intervention within this high genetic risk patient group.


INTRODUCTION
22q11.2 Deletion Syndrome (22q11.2DS)has been identified as one of the strongest risk factors for the development of neurodevelopmental and psychiatric conditions across the lifespan, including intellectual disability, attention deficit hyperactivity disorder (ADHD), anxiety and mood disorders, autism spectrum conditions, and is one of strongest known biological risk factors for schizophrenia (Malhotra & Sebat, 2012;Murphy et al., 1999;Niarchou et al., 2014;Rees et al., 2016;Sanders et al., 2015;Schneider et al., 2014).Clinical studies report over 60% of individuals who have been diagnosed with 22q11.2DS in a medical genetic clinic setting meet criteria for a psychiatric condition (Bertrán et al., 2018;Jonas et al., 2014).22q11.2DShas also been associated with psychiatric risk in genome-wide association studies (Malhotra & Sebat, 2012;Rees et al., 2014), and within population cohorts (Olsen et al., 2018).Although the effect size of associations between 22q11.2DS and psychiatric outcomes differs by study design and ascertainment strategy (Fiksinski, Schneider, et al., 2021), there is an overarching consensus across studies that it is associated with elevated psychiatric risk.
22q11.2DS affects 1 in 4000 live births and is considered the most frequent chromosomal microdeletion syndrome (McDonald-McGinn et al., 2015), yet there is still relatively low awareness of this condition in the clinical community.22q11.2DS is typically diagnosed within the first years of life, usually following referral to medical genetics clinics for a range of clinical features, particularly congenital abnormalities and developmental concerns (Cancrini et al., 2014).This provides unique opportunities for prospective deep phenotyping to identify transdiagnostic markers of risk (Baker & Vorstman, 2012) that manifest early in development and underlie later development of psychiatric conditions.This knowledge is needed to develop tailored services for prevention and early detection and management of serious mental illness.
Most studies of 22q11.2DS in childhood have focused on children aged 6 years and above, and these have established that divergent cognitive and psychopathology trajectories are present in late childhood and adolescence (Chawner et al., 2017;Chawner, Niarchou, et al., 2019;Morrison et al., 2020;Schneider et al., 2014;Tang & Gur, 2018).Few studies of psychopathology in children with 22q11.2DShave focused on earlier developmental periods.We are only aware of only two such studies (Klaassen et al., 2013;Kortanek et al., 2022).In this, 1.5 to 6-year-olds with 22q11.2DSwere screened with the Child Behaviour Checklist (CBCL) to investigate the presence of early behavioural and emotional problems (Klaassen et al., 2013), and it was found that 30% of children met clinical cutoffs for behavioural and emotional problems.Three studies have described a high prevalence of developmental concerns in preschool children with 22q11.2DS(Gerdes et al., 1999(Gerdes et al., , 2001;;Kortanek et al., 2022).54% of children with 22q11.2DShave significant developmental delay, and 80% show delays in language development (Gerdes et al., 1999(Gerdes et al., , 2001)).Although previous work highlights atypical early development, the phenotypic depth of these studies has been limited, and there is a need for further studies with wideranging assessments to capture a breadth of early indicators of neurodevelopmental and psychiatric liability.Furthermore, previous studies of early childhood in 22q11.2DShave not consistently included a group of unaffected controls to investigate the extent to which developmental features in 22q11.2DSare specific.Therefore, the impact of 22q11.2DS on early development relative to typically developing children has not been quantified.
An additional limitation of previous work has been a focus on categorical clinical cut-offs, rather than the dimensions that underlie early childhood impairment.The National Institute of Mental Health has developed a framework (Research Domain Criteria [RdoC]) which -rather than characterising individuals in terms of categorical diagnostic mental health disorders-uses a broader approach based on dimensional constructs that underlie these conditions.These RdoC framework dimensional measures are not simply continuous measures of categorical psychiatric diagnoses; instead, there is a focus on measures that index the underlying mechanisms and neurobiological bases of psychiatric conditions.Methodologies for assessing potential dimensional RdoC traits range from subjective questionnaire measures to objective assessment of traits via approaches including cognitive testing and experimental eye-tracking paradigms (National Institute of Mental Health, 2011).Previous studies of school-age children with 22q11.2DShave reported behavioural (Schneider et al., 2014), cognitive (Fiksinski, Bearden, et al., 2021)  independent of the presence of childhood neurodevelopmental and psychiatric diagnoses, including ADHD, autism, and anxiety (Niarchou et al., 2014), but the extent this applies to earlier developmental periods has not been investigated.
The RDoC approach offers increased opportunities to examine transdiagnostic liability (Doherty & Owen, 2014;Insel et al., 2010).This approach may be particularly informative in young children, as it can capture subthreshold liability not yet manifested as a clinical diagnosis but that may be indicative of later risk.Another benefit of dimensional measures is that they open the possibility of investigating whether subthreshold traits in relatives predict outcomes in children at risk.Variability in cognitive and psychiatric traits in 22q11.2DShas been found to be predictable, based on the cognitive and psychiatric profiles of relatives (Klaassen et al., 2014;Olszewski et al., 2014), highlighting the fact that factors beyond the deletion influence neurodevelopmental and psychiatric outcomes.However, the extent to which this applies to early development in 22q11.2DShas not previously been investigated, and therefore it is unclear whether phenotypic correlations between children and relatives emerge through development or if this is present from early development.
Here we present findings from young children aged 2-5 years old with 22q11.2DScompared to unaffected sibling controls.Children completed a broad phenotyping battery which included: (a) age appropriate dimensional measures (Hay et al., 2021)

Assessments
Parents and caregivers were asked to complete questionnaires in advance of the research assessment appointment, and interviews took place at the assessment.Children with 22q11.2DSand sibling controls within the same family were assessed on the same day, but in separate rooms by separate raters.

Developmental history
Developmental history was ascertained via caregiver report questionnaires and interviews.

Developmental milestones
The developmental milestones section of the Autism Diagnostic Interview-Revised (ADI-R) (Rutter et al., 2003) was administered with the caregiver to ascertain the age in months in which milestones were achieved for walking, toilet training, and language development.

Neurological health
The Epilepsy Screening Questionnaire (ESQ) (Ottman et al., 2010) was administered to screen for epilepsy diagnosis and seizure-like symptoms.

Dimensional phenotyping
Table 2 provides an overview of the range of dimensional phenotypic traits that were assessed and how they align to RDoC domains and psychiatric phenomenology.

Cognitive function
All children completed the Mullen Scales of Early Learning (Mullen, 1995) (Meeuwsen et al., 2019).Theory of Mind was assessed using tasks from Wellman and Liu's ( 2004) theory of mind scale (Wellman & Liu, 2004).

Oculomotor function
Children completed established eye-tracking paradigms to assess voluntary and spontaneous oculomotor behaviour, including tests of prosaccades, smooth pursuit and fixation (Barton et al., 2008;Holzman, 2000;Morita et al., 2020).Full details of eye-tracking methodology can be found in the Supporting Information S1.

Motor development
We used two measures of motor functioning: (a) The Preschool Developmental Coordination Questionnaire (Little DCDQ) (Rihtman et al., 2011), was used to screen for liability for Developmental Coordination Disorder (Wilson et al., 2000); (b) The motor sections of the Vineland Adaptive Behaviour Scales (Sparrow et al., 1984) (designed for children with intellectual disabilities) were used to provide an assessment of motor functioning, from which standardised scores for Overall Motor Functioning, Gross Motor Functioning, and Fine Motor Functioning were derived.

Sleep functioning
The Tayside Children's Sleep Questionnaire (TCSQ) was used to screen for disorders of initiating and maintaining sleep in children aged between 1 and 5 years (McGreavey et al., 2005).A symptom score can be derived from the TCSQ, as well as a categorical indicator of clinically relevant sleep problems.

Comparing developmental history between 22q11.2DS and controls
For each developmental milestone, as ascertained from the developmental interview, age was compared between children with 22q11.2DSand controls using linear mixed models that controlled for gender and age as fixed effects, and familial relatedness as a random effect (to account for the fact that some children with 22q11.2DSand controls came from the same family).Missing variables occurred as not all caregivers were able to retrospectively provide exact age in months for developmental milestones.For analysis of epilepsy-related variables, we conducted Fisher's exact tests for seizure presence due to 0 cell count for the controls.For analysis of presence of epilepsy symptoms, mixed effect models for binary outcomes failed to converge, so we opted for logistic regression models with gender and age as covariates.

Comparing developmental history between 22q11.2DS and controls
The following results were based on the parent report developmental

Contrasting 22q11.2DS and controls on dimensional and categorical measures of cognition, motor development, sleep function, and neurodevelopmental and psychiatric liability
Children with 22q11.2DSdiffered from controls on a range of cognitive, motor, oculomotor, sleep and neurodevelopmental and psychiatric traits analysed (Table 3; Figure 1), and the effect size of the differences (based on Hedges' g; Cohen, 1988) was large (≥0.8) for the majority of traits.Furthermore, the majority survived B-H FDR multiple testing correction, and remained significant in sensitivity analyses including household income as a covariate (Table S2).
Also for traits where population norm or community control data were available, children with 22q11.2DSscored significantly below the norm for the majority (28/30) of traits (Table S3, z-tests p < 0.001).For the majority (26/30) of traits the sibling controls did not differ significantly from population norms or community control data (Table S3).Although this indicates that sibling controls were broadly representative of population norms for traits including global cognitive ability, overall autistic traits and total score for behavioural and emotional problems (CBCL), it should be noted sibling controls did score significantly higher (higher score should be interpreted as  Table 4).The significance of these findings survived B-H FDR 0.05 correction for multiple testing.

Association of cognitive, motor and sleep markers with neurodevelopmental and psychiatric outcomes
Within children with 22q11.2DS,behavioural and emotional problems total score (CBCL) was significantly associated with developmental coordination score (r = 0.53, p = 0.002) and sleep problems (p < 0.001) (see Table 5).It should be noted that the behavioural and emotional problems total scores also correlated with global cognitive ability (r = 0.38, p = 0.037), but this did not survive multiple testing correction.

Influence of familial traits on childhood traits
Parental SRS T scores were lower, reflecting lower autistic traits, when  p-values were derived from logistic regression models were conducted with age and gender as covariates, except for those starred (*).* For some categorical variables where there was 0 cell count for controls, Fisher's exact test was used to derive p-values.impairments across a broad range of domains, including cognitive, motor, oculomotor, language, social, sleep and neurodevelopmental and psychiatric impairments, relative to non-carrier sibling controls and population norms.This emphasises that psychiatric liability manifests at an early age, many years before adolescence and young adulthood when psychiatric disorders typically emerge.High levels of phenotypic variability have been previously reported for older children and adults with 22q11.2DS(Chawner et al., 2021;Davies et al., 2020;Jacquemont et al., 2022).Here, we find that phenotypic variability is already present at a very early age.(Chawner et al., 2017).
We also identified that motor and sleep impairments in early childhood index neurodevelopmental and psychiatric outcomes.We  (Ching et al., 2020;Sun et al., 2020).Our findings for 22q11.2DScorroborate those for polygenic risk for schizophrenia, that is, that genetic risk impacts motor development from an early age many years before the onset of psychosis (Serdarevic et al., 2018).
Our study provides evidence that in addition to the effects of the 22q11.2deletion, familial traits influence the presence of autistic traits in early childhood in 22q11.2DS.This highlights that factors other than 22q11.2DSpredict risk; parental autism scores predicting scores in children with 22q11.2DScould reflect polygenic and other family-related factors (Jacquemont et al., 2022).Previous studies have found that polygenic scores for schizophrenia and intellectual intelligence were predictive of schizophrenia and intellectual in individuals with 22q11.2DS(Davies et al., 2020).
Although this study had the strength of including a control group, and assessing a broad range of dimensional phenotypes, the findings should be considered in light of a range of limitations.
Firstly, children were ascertained for the study on the basis of an existing genetic diagnosis, and this is likely to introduce ascertainment bias towards children with developmental delay as this is a common reason for referral to genetic testing.However, it should be noted, that despite this, considerable variability in outcomes was still seen within this cohort, emphasised by the cluster analysis findings that identified a subgroup of 22q11.2DSchildren at lower vulnerability with similar scores to control siblings.Comparison to population norms and community control data revealed that broadly, the scores of the sibling controls were representative of the general population, but is important to highlight there were a small number of traits whereby sibling controls did differ, highlighting the importance of considering a range of control groups when designing future studies.The differences between children with 22q11.2DS on neuropsychiatric measures could be partly explained by the presence of intellectual disability.In our study we find that although global cognitive ability correlated with autistic traits (r = 0.38) and CBCL total scores (r = 0.38), and explained 14% of the variance for each outcome, however these findings did not survive multiple testing correction.We cautiously conclude that global cognitive ability may play a role, but cannot fully explain the presence of neuropsychiatric problems in children with 22q11.2DS.This is consistent with findings in older children that find intellectual disability and psychiatric problems are independent consequences of 22q11.2DS(Niarchou et al., 2014).Although the sample size of the study is relatively small, study size is consistent with other early developmental deep phenotyping studies of rare genetic variants (Hogan et al., 2017;Kolesnik et al., 2017;McDonald et al., 2017).We have interpreted findings cautiously, and applied multiple testing corrections, nonetheless our findings should be regarded as exploratory and warrant replication.Our study size was sufficient to detect the large effect sizes conferred by 22q11.2DS on early development, which are likely to have clinical significance, but would not have been powered to detect more subtle differences.The cluster analysis findings should be regarded as exploratory due to sample size and an imbalance in group sample sizes.The transdiagnostic approach of the cluster analysis, as used in previous research (McDougal et al., 2020), was not intended to be used to make broad claims.
Rather, the purpose was to investigate whether children with 22q11.2DSand sibling controls clustered together or were distributed across different clusters.Future research is needed in larger samples to investigate the factor structure that underlies the range of cognitive and behavioural impairments present in the early childhood of 22q11.2DS.Large consortium approaches have been applied to investigating the later childhood and adult phenotypes of 22q11.2DS(Chawner, Owen, et al., 2019;Chawner et al., 2021;Schneider et al., 2014;Vorstman et al., 2015), but the same needs to be applied to studies of early development and across a range of risk loci.

CONCLUSION
22q11.2DS has a large and broad impact on early childhood across motor, social, language and cognitive development, and a range of transdiagnostic clinical risk indicators can be detected from an early age, years before adolescent and adult psychiatric problems may develop.There appears to be a group of young children with 22q11.2DSwho diverge from their 22q11.2DSpeers and control siblings and express higher levels of neurodevelopmental and psychiatric liability.Motor and sleep function appear to be markers of early neurodevelopmental and psychiatric liability in 22q11.2DSand thus may represent early targets for intervention.Overall, our findings highlight the need for future research that takes a developmental approach to understanding how genetic risk for psychiatric conditions manifests.Large cohorts of children with rare psychiatric variants need to be established from birth and followed longitudinally, complemented by studies that investigate the impact of rare psychiatric risk variants on early development in population cohorts.

�
and motor impairments (Cunningham et al., 2018), but the extent to which these covary in early development remains unclear.Intellectual impairments in school-age children with 22q11.2DShave been found to be Key points � 22q11.2deletion syndrome (22q11.2DS) is a genetic condition that greatly increases liability for neurodevelopmental and psychiatric outcomes.Examining the early developmental phenotype in 22q11.2DScould elucidate early transdiagnostic markers of childhood vulnerability.Young children with 22q11.2DSand their unaffected siblings were assessed using a range of dimensional measures informed by the Research Domain Criteria framework; including sleep functioning, cognitive, motor, reward valuation and social processes.� Although neuropsychiatric conditions typically emerge later in adolescence in 22q11.2DS,our study identified a range of risk indicators in early childhood, indicating scope for future studies of early risk mechanisms and early intervention in this at-risk patient group.
aligned to the RdoC framework, and (b) measures of neurodevelopmental and psychiatric liability from which both dimensional and categorical outcomes could be derived.Our specific aims were: (1) to assess the early developmental profiles of children with 22q11.2DSrelative to controls; (2) to investigate whether cognitive, motor and sleep traits were associated with neurodevelopmental and psychiatric liability, and (3) to investigate whether variability in early childhood outcomes is influenced by the phenotypes of relatives.
caregiver report questionnaire was administered from which the following scores were derived: (a) transdiagnostic domains including Internalising Problems (comprised of Emotionally Reactive, Anxious/Depressed, Somatic Complaints, Withdrawn domain scores), Externalising Problems (comprised of Attention Problems, and Aggressive Behaviour domain scores), and Stress Problems; (b) DSM-oriented scales including liability scales for Depressive Problems, Anxiety Problems, Attention Deficit/Hyperactivity Problems, and Oppositional Defiant Problems.Total Problems Score was also derived which represented overall neurodevelopmental and psychiatric liability.The Social Responsiveness Scale Second Edition (SRS-2) (Constantino & Gruber, 2012) provides a dimensional measure of autismrelated traits.The caregiver report version of the SRS was administered to ascertain autism traits in children.The following standardised scores were derived: SRS Total T-Score, Social Awareness T-score, Social Cognition T-score, Social Communication T-score, Social Motivation T-score, RRB (Restricted Interests and Repetitive Behaviour) T-Score.The adult report version of the SRS was administered to ascertain autism traits in biological mothers and fathers.
Rihtman et al., 2011;Wellman & Liu, 2004).Population norm data or community control data was not available for the eye-tracking tasks.Z-tests were conducted to investigate how the 22q11.2DSand sibling control groups performed relative to samples representative of the general population.Correlations between the dimensional traits were calculated to explore the relationships between traits.Exploratory K-means cluster analysis was conducted to categorise children with 22q11.2DSand controls into clusters based on phenotypic similarity.Previous neurodevelopmental research with a similar sample size has used exploratory cluster analysis to characterise transdiagnostic heterogeneity(McDougal et al., 2020).The following dimensional traits, adjusted for age and gender, were included in the cluster analysis: from the Mullen Scales of Early Learning (cognition); CBCL (neurodevelopmental and psychiatric traits; DSM-based and transdiagnostic), SRS (autistic traits), Vineland (motor functioning), DCDQ (motor coordination) and the Tayside questionnaire (sleep).The following summary scores were excluded from the cluster analysis due to being mathematically related to their subdomain scores which were already included in the analysis: SRS Total Score, Total CBCL Score, Internalising Problems CBCL Score, Externalising Problems CBCL Sore, Vineland Motor Standard Score and the Mullen General Cognitive Ability Score.Dimensional traits derived from the eyetracking tasks were not included in the cluster analysis due to missing data, as not all children engaged in the task and not all data passed strict quality control thresholds (see TableS1).The fviz_nbclust R package was used to identify the optimal number of clusters.Children with 22q11.2DSand controls were also compared on categorical variables based on established clinical screening thresholds, including Global Cognitive Ability (Mullen Early Learning Composite score ≤85 indicates below average ability), Vineland Motor functioning impairment (standardised score ≤85), Developmental Coordination liability (≤67 for boys, ≤68 for girls), SRS (standardised score ≥60), Total CBCL score and DSM (Anxiety, ADHD, Mood and ODD) subscales (clinical liability categories derived by CBCL software based on score, age and gender), and sleep problems (clinical threshold ≥8).When comparing prevalence of these clinical liability indicators between 22q11.2DS and controls, mixed effect models taking account of fixed effects of age, gender and relatedness as a random effect were conducted.For variables where the model failed to converge, logistic regression models were constructed with age and gender as covariates.For some categorical variables there was 0 cell count for controls, and thus Fisher's exact test was used to compare prevalence between groups where logistic regression was not appropriate.B-H FDR multiple testing correction value of 0.05 was applied to analyses.Association of cognitive, motor and sleep markers with neurodevelopmental and psychiatric outcomes in children with 22q11.2DSExploratory correlation analyses were conducted to investigate which cognitive and motor traits indexed neurodevelopmental and psychiatric outcomes, including autistic traits (SRS) and total score for Behavioural and Emotional Problems (CBCL).Trait scores were standardised into z-scores adjusted for age and gender before being included in correlation analyses.B-H FDR multiple testing correction value of 0.05 was applied to analyses.Influence of familial traits on childhood traitsWe investigated if the SRS score of children with 22q11.2DSwas influenced by biological parental SRS score, following the approach of a previous study(Moreno-De-Luca et al., 2015) that calculated intraclass correlation coefficients (ICC).Here, SRS ICC scores were calculated for 31 mother-child pairs and 22 father-child pairs.
Figure2and TableS4).Cluster analysis of dimensional traits (full list in methodology section) identified two groups of participants (Cluster 1 = 25 participants, controls = 11, 22q11.2DS= 14; Cluster 2 = 16 participants, all children with 22q11.2DS),which accounted for compared to those of children with 22q11.2DS; 31 mother-child pairs, maternal SRS mean = 48.0,22q11.2DSmean = 60.9, paired t-test p < 0.001; 22 father-child pairs, paternal SRS mean = 49.0,22q11.2DSF I G U R E 2 Correlation heatmap of dimensional cognitive and behavioural traits.CBCL, Child Behaviour Checklist; Little DCDQ, The Preschool Developmental Coordination Questionnaire; SRS, Social Responsiveness Scale Second Edition; TCSQ, The Tayside Children's Sleep Questionnaire.mean = 61.1,paired t-test p < 0.001.Total SRS score was highly and positively correlated between children with 22q11.2DSand mothers (ICC = 0.47, p = 0.046, n = 31), whereas there was no significant evidence for an association with paternal SRS score (ICC = 0.28, p = 0.22, n = 22).Though it should be noted, that fewer fathers were available to complete the SRS compared to mothers.DISCUSSIONThis study took a genetics-first approach(Lord & Veenstra- VanderWeele, 2016) to investigate how early child development is impacted by high genetic risk for neurodevelopmental and psychiatric outcomes.We found that young children with 22q11.2DSshow F I G U R E 3 Study participants clustered based on dimensional phenotypes.Dimension 1 (Transdiagnostic) and 2 (Affective-cognitive), represent the principal components that explained the most variability across the dimensional phenotypes.T A B L E 4 Categorical outcomes in children with 22q11.2DSand controls.
cannot infer direction of effect from this study and these findings are exploratory, but our results highlight motor development and sleep function as early markers of psychiatric risk in children with 22q11.2DS,and mirror findings in older children and adolescents with 22q11.2DSthat indicate that sleep and motor impairments index psychiatric risk (Cunningham et al., 2018; Moulding et al., 2020).There are several possible explanations as to why early motor development and sleep problems may index neurodevelopmental and psychiatric risk.First, these traits may directly lead to the development of neurodevelopmental and psychiatric impairment.Another possibility is that motor and sleep problems are a secondary consequence of early neurodevelopmental and psychiatric impairment.Thirdly, motor and sleep function could codevelop with neurodevelopmental and psychiatric liability, for example, as a consequence of the aberrant brain development seen in 22q11.2DS , including via NHS medical genetics clinics, the patient support group Max Appeal, and social media.Children were eligible if they were aged between 24 and 71 months and had a diagnosis of 22q11.2DSconfirmed by a medical genetics clinic; control siblings without 22q11.2DSwithin the same age range were also invited to take part.Exclusion criteria based on neu- Demographics of children with 22q11.2DSand controls.
Global Cognitive Ability refers to Mullen's Early Learning Composite score.Full details and references for the measures listed can be found in the methodology section.
(Hay et al., 2014(Hay et al., , 2021;;McGreavey et al., 2005;Meeuwsen et al., 2019;ative to the sibling controls (reference group), which were adjusted for the covariates of gender and age.Scores were constructed so that a negative score indicated atypical development.Linear mixed models were conducted to investigate developmental differences between children with 22q11.2DSandcontrols,wherebygenderand age were fixed effects and familial relatedness was included as random effect.Benjamini-HochbergConstantino & Gruber, 2012;Mullen, 1995; Sparrow, 2011), whilstfor others where possible we were able to use previously collected community control data (Tayside questionnaire, DCDQ, Cardiff Child Development Study tasks, Theory of Mind scale)(Hay et al., 2014(Hay et al., , 2021;;McGreavey et al., 2005;Meeuwsen et al., 2019;T A B L E 2 Overview of dimensional traits assessed. (Niarchou et al., 2014)velopmental traits and neurodevelopmental and psychiatric outcomes.Although our study was not longitudinal and can therefore not make claims about future risk, it is striking that in early childhood it is possible to identify a subgroup (16/30) who appear to be diverging in their development from other children with 22q11.2DSandcontrolsiblings, this proportion is in line with the prevalence of psychiatric conditions within school-age children with 22q11.2DS(54%)(Niarchouetal., 2014).Intervention strategies for . Our work highlights that many of the cognitive and psychopathology risk indicators described in adolescence are disrupted from early childhood.Previous research has indicated that cognitive trajectories of children with 22q11.2DSwholaterdevelopment psychosis start to diverge from age 11 onwards from T A B L E 5 Note: Bold number indicates the p-value survives B-H FDR 0.05 correction for multiple testing.NEURODEVELOPMENTAL DIMENSIONAL ASSESSMENT OF YOUNG CHILDREN children with 22q11.2DSwhodonot develop psychosis (Vorstmanet al., 2015).