Early Life Predictors of Child Development at Kindergarten: A Structural Equation Model using a Longitudinal Cohort

Sarah E Turner^1,2Stephanie Goguen^1,2Brenden Dufault³Teresa Mayer⁴Puishkumar J Mandhane^5,6Theo J Moraes⁷Stuart Turvey^8,9Elinor Simons^1,10,2Padmaja Subbarao^11,7Meghan B Azad^1,10,2*

• ¹Children's Hospital Research Institute of Manitoba (CHRIM), Winnipeg, Manitoba, Canada
• ²Department of Community Health Sciences, Max Rady College of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada
• ³Center for Health Care Innovation, University of Manitoba, Winnipeg, Canada
• ⁴General Child and Family Services Authority, Manitoba, Winnipeg, Canada
• ⁵Department of Pediatrics University of Alberta, Edmonton, Canada
• ⁶Faculty of Medicine UCSI University, Negeri Sembilan, Malaysia
• ⁷Department of Paediatrics, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
• ⁸British Columbia Children's Hospital, Vancouver, British Columbia, Canada
• ⁹Division of Allergy & Immunology, Department of Pediatrics, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada
• ¹⁰Department of Pediatrics and Child Health, University of Manitoba, Winnipeg, Manitoba, Canada
• ¹¹Department of Physiology, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada

Introduction: Early child development sets the stage for lifelong health. Identifying early life factors related to child development can help guide programs and policies to bolster child health and wellbeing. The objective of this research was to examine how a broad range of predictors, measured prenatally to the third year of life, are related to child development at kindergarten. Methods: We linked survey data from the Manitoba site of the CHILD Cohort Study with data from the Early Development Instrument (EDI) assessment, completed in kindergarten by the Manitoba public school system (n=442 children). The EDI measures five domains of development (ex. language, physical), scored to indicate the bottom 10% (i.e. 'vulnerable') of the population on one or more domains. Using structural equation modelling, we grouped 23 predictors of child development into six latent factors including prenatal exposures, child health and lifestyle, family stress, and socioeconomic status (SES). We examined the associations between each latent factor and EDI vulnerability.Results: Overall, 20.1% of children were vulnerable on one or more EDI domains. Higher family stress at 1 year and 3 years was related to a 0.20 (p-value ≤ 0.001) and 0.33 (p-value ≤ 0.001) standardized increase of EDI vulnerability. Higher socioeconomic statusSES was related to a -0.26 (p-value =0.01) standardized decrease of EDI vulnerability, and this link was partially mediated through family stress at three years (10.6% mediated). Prenatal exposures (e.g. maternal diet quality), as well as child health and lifestyle factors (e.g. weekday sleep) were not related to EDI vulnerability.Conclusions: Supporting parental mental health and throughout early life programs to reduce early life parenting stress to reduce , universal screening for early life stress, as well as targeting