Associations between clusters of early life risk factors and developmental vulnerability at age 5: a retrospective cohort study using population-wide linkage of administrative data in Tasmania, Australia

Objective: Early childhood is a critical time to address risk factors associated with developmental vulnerability. This study investigated the associations between clusters of early life risk factors and developmental vulnerability in children's first year of full-time school at age 5.

Design: A retrospective cohort study.

Setting: Population-wide linkage of administrative data records for children born in Tasmania, Australia in 2008-2010.

Participants: The cohort comprised 5440 children born in Tasmania in 2008-2010, with a Tasmanian 2015 Australian Early Development Census (AEDC) record and a Tasmanian Perinatal Collection record.

Outcome measure: The AEDC is a national measure of child development across five domains: physical health and well-being, social competence, emotional maturity, language and cognitive skills (school-based), and communication skills and general knowledge. Children who scored below the 10th percentile on one or more AEDC domains were classified as developmentally vulnerable. Children with special needs are not included in the AEDC results.

Results: Latent class analysis identified five clusters of risk factors: low risks (65% of children), sociodemographic and health behaviour risks (24%), teenage mother and sociodemographic risks (6%), birth risks (3%), and birth, sociodemographic and health behaviour risks (2%). In this sample population, 20% of children were classified as developmentally vulnerable, but the proportion varied substantially by latent class. Logistic regression showed increased odds of developmental vulnerability associated with sociodemographic and health behaviour risks (OR 2.26, 95% CI 1.91 to 2.68, p<0.001), teenage mother and sociodemographic risks (OR 2.01, 95% CI 1.50 to 2.69, p<0.001), and birth, sociodemographic and health behaviour risks (OR 3.29, 95% CI 2.10 to 5.16. p<0.001), but not birth risks (OR 1.34, 95% CI 0.88 to 2.03, p=0.1649), relative to the reference group.

Conclusions: The patterning of risks across the five groups invites consideration of multisectoral policies and services to address complex clusters of risk factors associated with developmental vulnerability.