Individual Submission Summary
Share...

Direct link:

Two-part model: Finding risk/protective factors for the prevalence and severity of adolescents’ externalizing behavior problems

Sat, October 6, 10:45am to 12:15pm, Doubletree Hilton, Room: Redrock

Abstract

This study aims to investigate how individual genetic sensitivity (5-HTTLPR), attachment, fathers’ incarceration, peer victimization, family instability, and socioeconomic conditions are associated with the prevalence and the severity of externalizing behavior problems in adolescence. In addition to family socioeconomic status (Åslund et al., 2013) and attachment (Zimmermann, Mohr, & Spangler, 2009), 5-HTTLPR is associated with adolescents’ externalizing behavior problems. However, there has been little research integrating significant relationship factors and including the roles of biological fathers for children in at-risk families.
Data from 1,282 children in the Fragile Families and Child Wellbeing study were analyzed. These children and their mothers participated in the surveys, observations, and saliva sample collection from childbirth to 15 years old. Using two-part models in STATA 14.2 (Belotti, Deb, Mannning, & Norton, 2015), we tested a probit model for the prevalence of adolescents’ externalizing behavior problems and a GLM model for the severity of problems among adolescents showing externalizing behavior problems. As predictors, we included children’s 5-HTTLPR, child gender, childhood attachment, mothers’ education, fathers’ incarceration, family instability, family poverty, household size, and peer victimization in each model.
As shown in Table 1, the two-part model results showed that having two short (SS) alleles on 5-HTTLPR and secure attachment in early childhood lessened the predicted probability of the prevalence of externalizing behavior problems in adolescence. However, having an incarcerated father or experiencing peer victimization increased the predicted probability of the prevalence. For the severity among adolescents who showed externalizing behavior problems, if children had never lived with their biological fathers or experienced highly unstable relationships with biological fathers, it increased the risk of more severe externalizing behavior problems in adolescence. A decrease in family socioeconomic resources, having an incarcerated father, and experiencing peer victimization also increased the risk of more severe externalizing behavior problems. Having SS alleles was not significantly associated with the severity. However, after adding the interaction between 5-HTTLPR and peer victimization (Table 2), two copies of a short allele on 5-HTTLPR moderated the effect of peer victimization on the severity of externalizing behavior problems.
The results partially support Zimmermann et al. (2009) by showing that the presence of two short alleles on 5-HTTLPR predicts adolescents’ externalizing behavior problems. However, this study shows protective effects of SS alleles. Unlike Zimmermann et al. (2009), the interaction between 5-HTTLPR and childhood attachment was not significant to predict the prevalence or the severity of adolescents’ externalizing behavior problems.

Authors