The Family Check-Up Intervention Attenuates the Link Between Polygenic Risk and Trajectories of Aggression across Middle Childhood

Childhood aggression is normative and typically declines with age, but when it persists it may be due to a combination of early environmental and genetic influences. A few early interventions have shown to moderate the effect... [ view full abstract ]

Childhood aggression is normative and typically declines with age, but when it persists it may be due to a combination of early environmental and genetic influences. A few early interventions have shown to moderate the effect of single genes on childhood externalizing behavior (Bakermans-Kranenburg & van IJzendoorn, 2015). However, no study has tested the buffering effect of an intervention on a polygenic risk score (PRS) in predicting trajectories of middle childhood aggression. The present study examined how polygenic risk for childhood aggression influenced trajectories of aggression across middle childhood, and differences based on intervention condition.

Participants were from a longitudinal randomized prevention trial (Family Check-Up; FCU) designed to improve parenting (n_intervention=252; n_control=250), and consisted of 49% female, 45% Caucasian, 31% African American, 12% Hispanic, 10% Biracial, 2% Other. Parents reported on their child’s aggression from ages 5-9 using Achenbach’s CBCL. SNPs for the current PRS were drawn from a meta-GWAS of childhood aggression (Pappa et al., 2016), and filtered using gene-set enrichment analyses to retain functional SNPs at a p < .05 cutoff. A latent growth curve model was estimated for aggression from 5 to 9 years of age, with the effect of the PRS included on intercept and slope, and PRS differences examined by intervention condition. Analyses controlled for child age, sex, ethnicity, demographic risk factors and ancestry using principal components analysis.

An initial unconstrained model (RMSEA = .06, CFI/TLI = .99) with decreasing slope of aggression (to reflect decreasing mean levels from ages 5-9) had significant mean and variance of the intercept and slope (ps < .001). Next, the PRS and covariates were entered as predictors of the intercept and slope. The PRS was significantly associated with the slope (B = -.32, p = .004). In a multigroup model split by control/intervention condition the PRS was significantly associated with the slope in the control group (B = -.48, p = .001) but not the intervention group (B= -.10, p = .60; X²difference (1) = 4.75, p = .029).

Results indicate the FCU attenuated the link between polygenic risk for aggression and trajectories of aggression from 5 to 9 years of age, although replication is needed. This research highlights the utility of employing a polygenic and functionally-informed approach to indexing genetic risk. Importantly, results suggest that early intervention is effective in reducing trajectories of problem behavior among children at genetic risk.

Pappa, I., St Pourcain, B., Benke, K., Cavadino, A., Hakulinen, C., Nivard, M. G., … Tiemeier, H. (2016). A genome-wide approach to children’s aggressive behavior: The EAGLE consortium. Neuropsychiatric Genetics, 171(5), 562–572

Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2015). The hidden efficacy of interventions: Gene×environment experiments from a differential susceptibility perspective. Annual Review of Psychology, 66, 381-409.