Abstract:

The risk assessment of violent recidivism is especially important for the evaluation of risk among prisoners. As aggression is a stable factor of prisoners’ violent recidivism, investigating their aggressive behavior can prevent and reduce risk of violent recidivism after parole or release, and contribute to social security and stable development. Studies have shown that aggressive behavior is influenced by childhood adversities and genetic susceptibility (such as MAOA-uVNTR low activity allele). However, most of the studies only focus on linear association, neglecting the interaction between factors and non-linear relationships. Moreover, the assessment of prisoners’ aggression ignored the subtypes of aggression, and was often conducted by questionnaires. These problems constrain effective prediction of aggression. In the current study, latent class model is applied to identify the classification of childhood adversities among male prisoners and ordinary adults. Using an experiment paradigm and questionnaires, as well as behavioral data as aggressive indicators, we further analyze the relationship between childhood adversity and aggression, and the moderation by MAOA-uVNTR, COMT Val158Met, and 5-HTTLPR polymorphism. The findings would help identifying the genetic indicators of highly aggressive individuals and detect susceptible individuals affected by childhood adverse experiences. Moreover, they provide theoretical and empirical reference for predicting violent behavior and building related intervention through training or medication.

Key words: childhood adversity, gene, aggression, latent class analysis