3.1.3. Contributions of Social Determinants of Health to Prediction of Risk for Later Outcomes

Due to the population-wide implementation of the index, the EDI data, when linked with other data sources, offer the opportunity to examine developmental outcomes at kindergarten in relation to later outcomes for otherwise "undiagnosed" populations—for mental health outcomes, academic outcomes, or both.

Two sets of studies, one from Canada and one from Australia, provide examples of this opportunity. Thomson et al. studied the mental health of over 35,000 kindergartenage children in British Columbia using EDI data. The study examined the patterns of children's emotional maturity and social competence (based on the subdomains of the EDI) and investigated the degree to which sociodemographic variables were related to these patterns [47]. Using latent profile analysis, six distinct social–emotional profile groups were found, with membership in the lowest functioning groups associated with being male, having English as a second language, and lower household income. In a subsequent study, children were followed up to age 14 using administrative health databases [48]. The latent socioemotional functioning profiles were applied once again and were found to be associated with early-onset mental health conditions. An examination of sociodemographic characteristics revealed that boys, children in households with unmarried parents, younger mothers, and those receiving subsidies were overrepresented in the lower socioemotional functioning groups.

These findings were reflected in Australian research [49,50]. In a study by Green and colleagues, four developmental profiles were identified using the EDI domains and subdomains that were hypothesized to present varied levels of risk for future development of mental health disorders [49]. The authors found that the odds of being in the risk groups were related to several SDOH (e.g., socioeconomic disadvantage, maltreatment) and non-SDOH (e.g., parental history of mental illness and criminal offending) variables. In a study by Piotrowska and colleagues [50] linking kindergarten data to educational, health, and protection records up to 11 years of age, researchers explored the context of transition from competence to vulnerability and found that only about 22% of children deemed as typically competent on the EDI transitioned to later vulnerability; 42% of those identified with a cluster of emotional vulnerabilities in kindergarten were also vulnerable later and 41% of children with cognitive vulnerabilities remained vulnerable. Demographic factors that have been shown to impact child development and mental health, such as parental mental illness, parental offending, and evidence of use of child protection services, were powerful determinants in influencing a child's transition between developmental profiles.
