NSW-CDS Publications


  • Green, M.J., Harris, F, Laurens, K.R., Tzoumakis, S., Dean, K., Brinkman, S., Chilvers, M., Sprague, T., Stevens, R., Carr, V.J. (2018). The New South Wales Child Development Study (NSW-CDS) – Wave 2 (Child age 13 years). International Journal of Epidemiology, 45(5), p1396-1397k, dyy115, doi: 10.1093/ije/dyy115.


This report provides a description of the NSW Child Development Study cohort, and what has been measured in the cohort, following the second record linkage (Wave 2) in 2016. The child cohort is comprised of 91,635 children, defined by inclusion in the NSW Australian Early Development Census (AEDC) of 2009 at school entry (n=87,037) and/or the investigator-led Middle Childhood Survey (MCS)conducted in 2015 (n=27,792). The MCS captured 26.6% of the original 2009 AEDC cohort, and also brought 4598 new children into the cohort; 23,194 children contributed records to both the AEDC and MCS assessments. Wave 2 brings together each child’s birth, mortality, health, academic achievement, school enrolment, school suspensions and expulsions, child protection and criminal records, linked with their mothers’ perinatal records, and with both parents’ mortality, health and criminal records, where available. This report describes the linkage rates for each of these record sets, as well detailed information about each record set, and characteristics for the child and their parents. Published findings from the study to date, its strengths and limitations and future directions, are also discussed. 

  • Tzoumakis, S., Carr, V.J., Dean, K., Laurens, K.R., Kariuki, M., Harris, F., Green, M.J. (2018). Prenatal maternal smoking, maternal offending, and offspring behavioural and cognitive outcomes in early childhood. Criminal Behaviour and Mental Health, doi: 10.1002/cbm.2089 (Published online 5 September 2018).


This study examined the associations between quantity of prenatal smoking and frequency of maternal offending and offspring behavioural and cognitive outcomes at age 5 years, while accounting for other prenatal and family risk factors. Prenatal smoking and frequent maternal offending (comprising two or more offences) were associated with all developmental vulnerabilities examined, even after adjusting for other familial and prenatal risk factors. Population attributable fractions (PAFs) for prenatal smoking ranged from 5.3% to 15.8% and PAFs for maternal offending ranged from 3.4% to 11.8% across the offspring outcomes. These findings suggest prevention and intervention efforts during prenatal periods for vulnerable families may have benefits for both behavioural and cognitive problems in early childhood. 

  • Bore, M.* / Laurens, K.R.*, Hobbs, M.J., Green, M.J., Tzoumakis, S., Harris, F., Carr, V.J. (2018). Item Response Theory analysis of the Big Five Questionnaire for Children Short-Form (BFC-SF): A self-report measure of personality in children aged 11-12 years. Journal of Personality Disorders, In Press. (*shared lead authorship)

In press; link to publication will be uploaded as soon as possible

Brief self-report questionnaire measures of the Big Five personality dimensions in children are needed for use in large population studies, for which longer instruments are not suited. The Big Five traits describe individual differences in personality within the population, with the expression of each trait in an individual varying along a dimension. These Big Five traits include: Agreeableness (generous, honest, modest vs. selfish, aggressive, arrogant), Conscientiousness (dependable, achievement-driven, constrained vs. disorderly, laid-back, unambitious), Neuroticism (worried, anxious, sad, vs. emotionally stable, calm), Extraversion (warm, outgoing, cheerful vs. introverted, reserved, solitary, sombre), and Openness to experience (curious, exploratory vs. rigid, practical, traditional). This study established the psychometric properties (reliability, validity) of a brief self-report measure of the Big Five personality dimensions in children, using data from 27,415 children in the NSW Child Development Study cohort. The psychometric properties were examined using Item Response Theory methods, exploratory and confirmatory factor analyses. The findings demonstrate the suitability of the brief, 20-item short-form measure of the Big Five personality dimensions in children for administration online in large population-based studies. The questionnaire could be useful in research examining how personality traits in childhood might act as risk or protective factors for later adolescent and adult health and well-being outcomes, and thereby inform public health policy and practices that promote mental health and prevent mental illness.

  • Green, M.J., Kariuki, M., Dean, K., Laurens, K.R., Tzoumakis, S., Harris, F., Carr, V.J. (2018). Childhood developmental vulnerabilities associated with early life exposure to infectious and noninfectious diseases and maternal mental illness. Journal of Child Psychology and Psychiatry, 59(7), 801-810; doi: 10.1111/jcpp.12856 (Published online 26 December 2017).


 This study investigated the effects of in utero exposure to maternal infection and noninfectious diseases during pregnancy on offspring developmental vulnerabilities at age 5 years, in a population sample of 66,045 children drawn from the NSW Child Development Study (NSW-CDS) cohort. Maternal infectious diseases during pregnancy and early childhood infection conferred the largest associations with developmental vulnerabilities at age 5 years; maternal noninfectious illness during pregnancy also retained small but significant associations with developmental vulnerabilities even when adjusted for other physical and mental illness exposures and covariates known to be associated with early childhood development. Among all exposures examined, maternal mental illness first diagnosed prior to childbirth conferred the greatest odds of developmental vulnerability at age 5 years.  Prenatal exposure to infectious or noninfectious diseases appear to influence early childhood physical, social, emotional and cognitive developmental vulnerabilities that may represent intermediate phenotypes for subsequent mental disorders.

  • Dean , K., Green, M.J., Laurens, K.R., Kariuki, M., Tzoumakis, S., Sprague, T., Lenroot, R.,  and Carr, V.J. (2018). The impact of parental mental illness across the full diagnostic spectrum on externalising and internalising vulnerabilities in young offspring. Psychological Medicine; doi: 10.1017/S0033291717003786 (Published online 14 January 2018).


 This study examined patterns of association between externalising and internalising vulnerabilities in early childhood and parental mental disorder, in a population sample of 69,116 children drawn from the NSW Child Development Study (NSW-CDS) cohort. Measures of externalising and internalising vulnerability, obtained via the Australian Early Development Census, were linked to administrative health datasets to determine parental psychiatric diagnostic status. Parental mental illness, across diagnostic categories, was associated with all child externalising and internalising domains of vulnerability. These findings have important implications for informing early identification and intervention strategies in high-risk offspring and for research into the causes of mental illness.

  • Green, M.J., Tzoumakis, S., Laurens, K.R., Dean, K., Kariuki, M., Harris, F., O’Reilly, N., Chilvers, M., Brinkman, S.A., Carr, V.J. (2017). Latent profiles of early developmental vulnerabilities in a NSW child population at age 5 years. Australian & New Zealand Journal of Psychiatry; doi: 10.1177/0004867417740208 (Published online 6 November 2017).


This study determined classes of children among the general population who may be at risk for later mental disorder on the basis of early patterns of developmental functioning, and the associations of each risk-class with exposure to childhood maltreatment, parental mental illness, parental offending, and perinatal adversities, and other demographic factors. Data from 16 subdomains of the Australian Early Development Census (AEDC) were used to determine classes of children with shared patterns of developmental vulnerabilities in a population sample of 67,353 children drawn from the New South Wales Child Development Study (NSW-CDS) cohort. Four classes were identified, reflecting putative risk states for mental disorders: (1) disrespectful and aggressive/hyperactive behaviour, labelled ‘misconduct risk’ (6.5%); (2) ‘pervasive risk’ (4.0%); (3) ‘mild generalised risk’ (11.6%); and (4) ‘no risk’ (77.9%). The odds of membership in putative risk groups (relative to the no risk group) were greater among children from backgrounds of child maltreatment, parental history of mental illness, parental history of criminal offending, socioeconomic disadvantage, and perinatal adversities.  

  • Dix, K.L., Green, M.J., Tzoumakis, S., Dean, K., Harris, F., Carr, V.J., Laurens, K.R. (2018) The Survey of School Promotion of Emotional and Social Health (SSPESH): A brief measure of the implementation of whole-school mental health promotion. School Mental Health, doi: 10.1007/s12310-018-9280-5 (Published Online 28 June  2018).


This paper describes the development of a brief school-based instrument designed for use in research and in educational practice. The Survey of School Promotion of Emotional and Social Health (SSPESH)measures the extent to which a school has implemented policies and practices in four health-promoting domains: (a) creating a positive school community, (b) teaching social and emotional skills, (c) engaging the parent community, and (d) supporting students experiencing mental health difficulties. The questionnaire can be used to differentiate high, moderate, and low implementation of whole-school policies and practices that promote the emotional and social health of students. To develop the instrument, responses were gathered via an online survey of Principals in almost 600 Australian primary schools in New South Wales during 2015. The paper describes the preliminary psychometric properties of the instrument (reliability, validity), and the development of the implementation index using Latent Class Analysis methods. This brief 13-item survey can be used by researchers and evaluators for comparative and multivariate analyses. School leadership may find it useful as a brief tool to guide the identification of target areas for whole-school improvement across the four important health-promoting domains.

  • Harris, F.,* Smith, M.,* Laurens, K.R., Green, M.J., Tzoumakis, S., Kariuki, M., Carr, V.J. (2018). Validation of a two-factor model of the Best Start Kindergarten Assessment of literacy and numeracy. Australian Journal of Education, 62(1), 36-48, doi: 10.1177/0004944118761381.


This study examined the structure of the Best Start Kindergarten Assessment (Best Start) of literacy and numeracy, in a general population sample of 37,734 children drawn from the NSW Child Development Study (NSW-CDS) cohort. Analyses conducted on the 11 Best Start scales supported a two-factor structure underpinning literacy and numeracy attainment. This two-factor measure of attainment at school entry is comparable in structure to standardised measures of literacy and numeracy administered in later school years, and may thus facilitate research examining pathways of academic performance over time.


  • Green, M.J., Tzoumakis, S., McIntyre, B., Kariuki, M., Laurens, K.R., Dean, K., Chilvers, M., Harris, F., Butler, M., Brinkman, S.A., Carr, V.J. (2017). Childhood maltreatment and early developmental vulnerabilities at age 5 years. Child Development, Aug 2017. doi: 10.1111/cdev.12928. 


 The study focused on the effects of early life maltreatment on early childhood developmental outcomes, including examination of the effects of multiple maltreatment types and the timing of maltreatment in a large population sample. We specifically examined associations between the type, and timing of childhood maltreatment and early developmental vulnerabilities in a population sample of 68,459 children (34,562 males) drawn from the New South Wales Child Development Study (NSW-CDS) cohort, using linked administrative data for children and their parents (collected from 2001-2009). The mean age of the child sample was 5.62 years (SD=0.37), and 2,135 children had already been exposed to (substantiated) maltreatment according to child protection records. Exposure to any type of childhood maltreatment was associated with pervasive developmental vulnerability at age 5 years. Children exposed to two or more maltreatment types, and with first maltreatment reported after 3 years of age, showed greater likelihood of vulnerability on multiple developmental domains relative to non-maltreated children. These patterns of association were found in the context of other important risk factors for early childhood developmental vulnerabilities that have been implicated in previous studies (e.g., male sex, maternal smoking during pregnancy). 

  • Tzoumakis, S, Dean, K, Green, MJ, Zheng, C, Kariuki, M, Harris, F, Carr, VJ, & Laurens, KR. The impact of parental offending on offspring aggression in early childhood: A population-based record linkage study. Soc Psychiatry Psychiatr Epidemiol, 2017, 54(4), 445-455. Published online 13 January 2017: doi: 10.1007/s00127-017-1347-3   


This study examined the impact of parental criminal offending, both paternal and maternal, on offspring aggression at age 5 years. Parental mental illness, child’s sex, and socioeconomic disadvantage were also accounted for in the analyses. Offspring aggression was assessed using the Aggressive Behaviour subdomain of the Australian Early Development Census (AEDC), a teacher-reported cross-sectional survey on child development. Parental criminal offending was obtained from NSW Bureau of Crime Statistics data and parental mental health was obtained from NSW Admitted Patients and Mental Health Ambulatory data. Parental history of violent and frequent offending increased the risk of high levels of aggression in offspring during early childhood, highlighting the need for intervention with affected families during this key developmental period.

  • Laurens, K.R., Tzoumakis, S., Kariuki, M., Green, M.J., Hamde, M., Harris, F., Carr, V.J., Dean, K. (2017). Pervasive influence of maternal and paternal criminal offending on early childhood development: A population data linkage study. Psychological Medicine, 47(5), 889-901.


This report examined the impact of parental criminal offending on developmental outcomes at age 5 years. Data from the Australian Early Development Census, a teacher-reported assessment of childhood development across five key domains, was linked with maternal and paternal offending histories obtained from official records. Children whose parents had a history of criminal offending were at significantly greater risk of developmental vulnerability on each of the five domains. The risk of vulnerability increased further when both parents offended, when mothers rather than fathers offended, and when the offending was of a violent nature. For all types of offending, children were more likely to be vulnerable on multiple domains than on a single domain. These findings indicate a widespread impact of parental offending on early childhood development and highlight the need to support affected families in order to improve outcomes.


  • Carr, V.J., Harris, F., Raudino, A., Luo, L., Kariuki, M., Liu, E., Tzoumakis, S., Smith, M., Holbrook, A., Bore, M., Brinkman, S.A., Lenroot, R.K., Dix, K., Dean, K., Laurens, K.R., Green, M.J. (2016). Cohort Profile: The New South Wales Child Development Study (NSW-CDS) – An Australian multi-agency, multi-generational, longitudinal record linkage study. BMJ Open, 6:e009023doi:10.1136/bmjopen-2015-009023.


This report provides a description of the characteristics of children who make up the NSW Child Development Study cohort. This group comprises 87,000 children who commenced their formal school education in 2009, and for whom class teachers completed the Australian Early Development Census (AEDC). Rates for the successful linkage of the AEDC records with other administrative data for the children and their parents are provided within this publication, and information about the demographic characteristics (e.g., sex, geographic and socioeconomic distributions), as well as the representativeness of this group to a comparable national population. The strengths and limitations of the study are also discussed.

  • Laurens K.R., Tzoumakis S., Dean K., Brinkman S.A., Bore, M., Lenroot, R.K., Smith, M., Holbrook, A., Robinson K.M., Stevens, R., Harris, F., Carr, V.J., Green, M.J (2017). The 2015 Middle Childhood Survey (MCS) of mental health and well-being at age 11 years in an Australian population cohort. BMJ Open 2017;7:e016244. doi:10.1136/bmjopen-2017-016244.


This report provides a description of responses to the Middle Childhood Survey (MCS), a computerised self-report assessment of children’s mental health and well-being completed by 27 808 children aged 11-12 years in New South Wales (NSW) schools in 2015. The MCS sample includes 85.9% of students enrolled at the 829 schools who participated in the MCS (35% of eligible schools in NSW), and these participating schools and children are representative of the NSW population. The MCS measured Social Integration, Prosocial Behaviour, Peer Relationship Problems, Supportive Relationships (at Home, School and in the Community), Empathy, Emotional Symptoms, Conduct Problems, Aggression, Attention, Inhibitory Control, Hyperactivity-Inattention, Total Difficulties (internalising and externalising psychopathology), Perceptual Sensitivity, Psychotic-Like Experiences, Personality, Self-esteem, Daytime Sleepiness and Connection to Nature. This report presents response distributions on each of these indices, indicating the range of competencies and vulnerabilities in the population. 

  • Kariuki, M., Raudino, A., Green, M.J., Laurens, K.R., Dean, K., Brinkman, S.A., Lenroot, R.K, Liu, E., Harris, F., Luo, L., Carr, V.J(2016). Hospital admission for infection during early childhood influences developmental vulnerabilities at age 5 years. Journal of Paediatrics and Child Health, 52(9), 882-888.

https://www.ncbi.nlm.nih.gov/labs/articles/27439883/ (Abstract)

This study examined the relationship between early childhood infections requiring hospitalisation and developmental vulnerability at age 5 years, as assessed using the Australian Early Development Census (AEDC). The AEDC provides an index of each child’s level of function on five domains, including Physical Health and Wellbeing, Social Competence, Emotional Maturity, Language and Cognitive Skills, Communication Skills and General Knowledge. Hospital admissions for infections were associated with vulnerability on all five developmental domains. This suggests that severe infections during early childhood may pose a risk to development by school-age.

  • Matheson, S.L., Kariuki, M., Harris, F., Green, M.J., Tarren-Sweeney, M., Dean, K., Tzoumakis, S., Brinkman, S., Chilvers, M., Sprague, T., Carr, V.J., Laurens, K.R. (2016). Effects of maltreatment and parental schizophrenia spectrum disorders on early childhood social-emotional functioning: a population record linkage study. Epidemiology and Psychiatric Sciences. doi: 10.1017/S204579601600055X


This study examined the associations between early childhood maltreatment (age 0-5 years) and childhood social and emotional functioning at approximately age 5 years. This was achieved by linking Family and Community Services Data, parental Mental Health Ambulatory and Admitted Patients Data and Australian Early Development Census Data.  Medium-sized associations were found between maltreatment and poor social competency, aggressive behaviour, and hyperactive/inattentive behaviour, and small associations were found between maltreatment and poor prosocial/helping and anxious/fearful behaviour. These associations overshadowed the impact of having a parent with schizophrenia spectrum disorders. The results suggest childhood maltreatment may adversely impact on early childhood social and emotional functioning to a greater extent than parental schizophrenia and related disorders. 


  • Laurens, K.R., Luo, L., Matheson, S.L., Carr, V.J., Raudino, A., Harris, F., Green, M.J. (2015). Common or distinct pathways to psychosis? A systematic review of evidence from prospective studies for developmental risk factors and antecedents of the schizophrenia spectrum disorders and affective psychoses. BMC Psychiatry.


This scientific review provides a summary of the available evidence relating to risk factors for the development of schizophrenia spectrum disorders (e.g., schizophrenia, schizoaffective disorder), and affective psychoses (e.g., bipolar disorder). A total of 127 papers were reviewed for the purpose of evaluating the evidence for various risk factors as potentially conferring risk for psychoses, and to highlight gaps in the evidence base.  Factors which conferred the greatest risk, or were observed most consistently to confer risk to both affective and non-affective (schizophrenia spectrum) psychoses, included: obstetric complications, maternal illness during pregnancy (especially infections), other maternal physical factors, negative family emotional environment, psychopathology and psychotic symptoms, and cognitive and motor dysfunctions. 


  • Sara, G., Luo, L., Carr, V.J., Raudino, A., Green, M.J., Laurens, K.L., Dean, K., Cohen, M. Burgess, P., Morgan, V.A. (2014). Comparing algorithms for deriving psychosis diagnoses from longitudinal administrative clinical records. Social Psychiatry and Psychiatric Epidemiology, 49(11), 1729-37.


This study describes a method that we have tested to determine the performance of four different methods for assigning a single diagnosis (for a psychotic disorder) from health records in which multiple diagnoses can be recorded for a single individual over time. Administrative health records are valuable research tools but the methods to extract accurate mental health diagnoses can be challenging because of the number of admissions for any given individual. The methods tested in this report include testing the validity of using: ‘any’ diagnosis, ‘most recent’ diagnosis, ‘most frequent’ (modal) diagnosis or ‘hierarchy’ (where a diagnostic hierarchy was applied). Overall agreement between administrative and reference diagnoses was modest with different methods producing results of varying accuracy. This illustrates that the choice of methods used to summarise health records can impact the accuracy of diagnosis and hence is an important consideration for study methods. 

  • Moore, E.A., Harris, F., Laurens, K.R., Green, M.J., Brinkman, S., Lenroot, R.K., Carr, V.J. (2014). Birth outcomes and academic achievement in childhood: A population record linkage study. Journal of Early Childhood Research, 12(3), 234–250.


This scientific report examined the association between early life risk factors and academic achievement at Kindergarten (approx. age 5) and in Grade 3 (approx. age 8). This was achieved by linking the NSW Midwives data collection with the Best Start Kindergarten Assessment. Increasing maternal age and lack of maternal prenatal smoking were associated with improved academic performance. The results suggest that programs that target children with less developed academic skills during the first year of school may improve subsequent results.