Role of maternal age at birth in child development among Indigenous and non-Indigenous Australian children in their first school year: a population-based cohort study.

BACKGROUND

Indigenous Australian children are twice as likely to score poorly on developmental outcomes at age 5 years than their non-Indigenous peers. Indigenous children are also more likely to be born to younger mothers. We aimed to quantify the relationship between maternal age at childbirth and early childhood development outcomes in Indigenous and non-Indigenous children.

METHODS

In this population-based, retrospective cohort study, we used data from the Australian Early Development Census (AEDC) that were probabilistically linked by the New South Wales (NSW) Centre for Health Record Linkage to several NSW administrative datasets, including the Perinatal Data Collection, the Register of Births, Deaths and Marriages (for birth registrations), the Admitted Patient Data Collection, and public school enrolment records, as part of the Seeding Success study. The resulting data resource comprises a cohort of 166 278 children born in NSW whose first year of school was reported in a 2009 or 2012 AEDC record (which were the years of AEDC data available at the time of data linkage). The primary outcome was the aggregate outcome of developmental vulnerability (scores in the bottom decile, according to the 2009 benchmark, on one or more of the five AEDC domains, which include physical, social, emotional, language and cognitive, and communication development). This outcome was measured in singleton children without special needs recorded on the AEDC, in those with available developmental data. As a secondary outcome analysis, we also repeated the main analyses on the outcome of developmental vulnerability on the individual domains. We estimated the absolute risk of developmental vulnerability by maternal age in Indigenous and non-Indigenous populations, and we also estimated the risk difference and relative risk between Indigenous and non-Indigenous children by use of modified Poisson regression.

FINDINGS

Of 166 278 children in the cohort, 107 666 (64·8%) children were enrolled in a public school in NSW in 2009 or 2012, of whom 7994 (7·4%) children were Indigenous (ie, they, or either parent, were recorded as Aboriginal or Torres Strait Islander on one or more birth records) and 99 672 (92·6%) children were not Indigenous. After exclusions, the final study population included 99 530 children (7206 [7·2%] Indigenous and 92 324 [92·8%] non-Indigenous). Of those for whom developmental outcome data were available, 2581 (35·9%) of 7180 Indigenous children and 18 071 (19·7%) of 91 835 non-Indigenous children were developmentally vulnerable on one domain or more. The risk of developmental vulnerability decreased with maternal ages between 15 and 39 years, but the decrease in risk with maternal age was significantly steeper in non-Indigenous than Indigenous children.

INTERPRETATION

Developmental vulnerability is most common in Indigenous and non-Indigenous children born to young mothers; however, Indigenous children have an increased risk of this outcome across most of the maternal age range. Policies that improve the socioeconomic circumstances of Indigenous children and families could promote better developmental outcomes among Indigenous children. Culturally appropriate support for Indigenous children, including those born to young mothers and disadvantaged families, could also reduce early childhood developmental inequalities.

FUNDING

The Australian National Health and Medical Research Council, Manitoba Centre for Health Policy.