As possible mediators in the association between childhood family income and self-harm, we included psychiatric disorders, substance abuse, self-harm, violent criminality and victimization, out-of-home placements, not being in education, employment or training (NEET) and school performance in adolescence (age 13–17).
We used the same strategy as with the parents to measure psychiatric disorders, substance abuse and violent criminality among the adolescents. Violent victimization in adolescence was derived from the same health care data (ICD10, Supplementary file 1). We also included a measure of self-harm in adolescence as a mediator, since history of self-harm is a strong predictor of subsequent self-harm episodes. Adolescent self-harm was measured in a similar manner as self-harm in young adulthood, excluding suicides (the analyses are restricted to children who are alive at the age of 18). We used binary indicators (ever during adolescence/no) for these measures.
NEET was defined as being unemployed or outside the labour force, but not retired or studying. The information was derived from Statistics Finland's data on main type of activity at the end of each year. A binary indicator was used in the analyses. School performance at the end of compulsory schooling was defined using information on grade point averages (GPAs) from Statistics Finland. We divided GPAs into quartiles. Since the GPA is registered only if the adolescent has applied for secondary school, usually at the age of 16, there were individuals with missing data (3%). We included these children in the lowest quartile since it is likely that – in the Finnish context where applying for secondary school is nearly universal and heavily encouraged during the last years of comprehensive school – they have a low GPA.
Lastly, all of the children that had any record of placement at age 13–17 in the child welfare register were defined as being placed out of home. In contrast to placements during childhood, out-of-home placements in adolescence are usually due to adolescent behavioural problems.
Do you have any questions about this protocol?
Post your question to gather feedback from the community. We will also invite the authors of this article to respond.