*3.2. Studying Social Determinants of Health among Special Populations of Children*

Hard-to-reach, vulnerable populations tend to be under-represented in research. As Brownell and colleagues reported in 2004, children living in lower socioeconomic (SES) neighborhoods tend to be less represented in educational data than those in higher SES neighborhoods [52]. In their analyses of Grade 3 standardized test outcomes, they found that greater percentages of children from lower SES neighborhoods either did not complete the provincial standardized tests, received an exemption from writing them, or were absent during the time the test was being written [52]. Due to the population-level reach of the EDI, it has been possible to examine associations between SDOH and developmental outcomes in a number of different special populations of children. In this section, we will focus on research involving immigrant and refugee children, children with health disorders, and children who experience maltreatment or who are placed in out-of-home care.

#### 3.2.1. Immigrant and Refugee Children

Immigrant and refugee children represent a socially, culturally, and economically diverse group, which in Canada is a growing percentage of the population. To date, the literature on child development outcomes of immigrant and refugee children tends to be sample-based and relies on parent reports, which while an important source of data on children, may not provide a representative picture, as families who do not speak the study language fluently are often excluded and there may be mistrust towards researchers. Recently, a group of Canadian researchers started examining the associations between the SDOH and developmental outcomes using EDI data linked with a range of other datasets. For example, guided by Bronfenbrenner's bioecological model [44,53], Milbrath and Guhn [54] examined the relationship between immigrant children's cultural background, neighborhood-level socioeconomic factors and cultural composition, and their developmental outcomes. Their study used EDI data linked with administrative immigration records and census data to examine the effects of family and neighborhood poverty, neighborhood cultural density (in terms of being similar or not to the child's culture), and immigrant generational status on children's developmental health at school entry among Cantonese, Mandarin, Punjabi, and Filipino children in comparison to non-immigrant, English-speaking children. In line with previous studies, they found a negative association between family and neighborhood socioeconomic disadvantage and children's EDI scores. They also found differences in the associations between a neighborhood's cultural diversity and children's developmental outcomes based on neighborhood SES indicators and children's cultural backgrounds, with Mandarin-speaking children having lower developmental outcomes in neighborhoods with greater cultural density and Punjabi-speaking children having better developmental outcomes in poorer neighborhoods with greater cultural density.

Another Canadian study by Gagné and colleagues [55] investigated the relationships between income and literacy and numeracy trajectories from kindergarten to Grade 7 for various groups of migrant children living in the Canadian province of British Columbia. They examined the three official categories of migrant children: economic, family, and refugee categories. They found that similarly to non-migrant children, lower income was associated with lower literacy and numeracy trajectories in all but one group of migrant children. Migrant children who were in the high-achieving economic class group were less impacted by low income. Gagné et al. [55] found that parental education levels and children's abilities in English predicted high literacy and numeracy trajectories, despite low income.

#### 3.2.2. Children with Health Disorders

Until recently, Canada has lacked nationally representative data pertaining to social indicators of young children's developmental health, especially for those with health disorders. The ability to link EDI data with other datasets has allowed researchers to conduct studies on children with health disorders that were not possible before, either because of non-representative samples or because of a lack of data on certain key variables. Here, we will describe some studies from Canada and Australia that have examined SDOH in kindergarten children with health disorders.

Using pan-Canadian EDI data linked to a custom,-built neighborhood-level SES index [41], Zeraatkar and colleagues [56] examined the relationships between neighborhoodlevel SES and developmental health in children with disabilities, as identified in the EDI. Their results showed that all developmental domains were positively correlated with neighborhood-level SES, with the strongest relationship evident in the language and cognitive development domain. This association had already been noted in typically developing children (e.g., [41]), however this was the first Canadian population-level study to examine this link in children with disabilities. Relatedly, in Australia, O'Connor and colleagues [57] found a link between neighborhood-level SES and the odds of having an established or emerging special health-care need, with children living in the most disadvantaged neighborhoods having the highest odds of having a special health-care need.

Other studies have focused on specific health disorders, such as autism spectrum disorder [58–60], fetal alcohol spectrum disorder (FASD) [61,62], and unaddressed dental needs in kindergarten [63]. These studies consistently demonstrated the relationships between children's diagnoses, health needs, and SDOH, such as indicators of socioeconomic status at the neighborhood level.
