Child maltreatment consists of any acts of commission or omission by a parent or other caregiver that result in harm, potential for harm to a child (0–18 years of age) even if the harm is unintentional (Gilbert et al. Reference Gilbert, Widom, Browne, Fergusson, Webb and Janson2009), and that can take the form of neglect, emotional maltreatment, physical abuse or sexual abuse (Barnett et al. Reference Barnett, Manly, Cicchetti, Cicchetti and Toth1993). Many studies underline a strong association between child maltreatment and its immediate and long-term psychopathological consequences (Wolfe et al. Reference Wolfe, Scott, Wekerle and Pittman2001; Bot et al. Reference Bot, de Leeuw den Bouter and Adriaanse2011); however, clinical evidence points out that not all subjects exposed to maltreatment will develop psychopathological symptoms; the variability in children's responses suggest that this heterogeneity may be within the sphere of genetic influence (Moffitt et al. Reference Moffitt, Caspi and Rutter2006). Caspi et al. (Reference Caspi, McClay, Moffitt, Mill, Martin, Craig, Taylor and Poulton2002) (Table 1) were the first to report on a G × E interaction in child maltreatment. They studied a large sample of male children from birth to adulthood and found that a functional polymorphism in the gene encoding the neurotransmitter-metabolizing enzyme monoamine oxidase A (MAOA) moderates the effect of maltreatment: individuals who were carriers of the low-activity allele, but not of the high-activity allele, were at an increased risk of antisocial behaviour disorders following maltreatment. After this work many researchers have adopted a multiple level of analysis over the course of the past few decades (Cicchetti & Blender, Reference Cicchetti and Blender2004). Recently, Fergusson et al. (Reference Fergusson, Boden, Horwood, Miller and Kennedy2011) (Table 1) have replicated this result using a 30-year longitudinal study with a sample composed by male subjects, considering only physical and sexual abuse. Furthermore, Derringer et al. (Reference Derringer, Krueger, Irons and Iacono2010) (Table 1) have extended the moderating role of MAOA to conduct disorder symptoms in a sample of same-sex twins for sexually abused subjects. Similar results were found by Aslund et al. (Reference Aslund, Nordquist, Comasco, Leppert, Oreland and Nilsson2011) and Nilsson et al. (Reference Nilsson, Comasco, Aslund, Nordquist, Leppert and Oreland2011) (Table 1) examining, respectively, delinquent behaviour and high alcohol consumption. MAOA polymorphism was also found to enhance depressive symptoms but only among extensively maltreated youth (i.e. three or four maltreatment subtypes) (Cicchetti et al. Reference Cicchetti, Rogosch and Sturge-Apple2007). Subsequent studies on G × E interaction in child maltreatment have focused their attention on several other genetic factors, such as serotonin-transporter-linked polymorphic region (5-HTTLPR), corticotropin releasing hormone receptor 1 (CRHR1) and brain-derived neurotrophic factor (BDNF). 5-HTTLPR polymorphisms were found to moderate the effect of maltreatment both on externalizing and internalizing symptoms. In a longitudinal study, an interaction between maltreatment and 5-HTTLPR on antisocial behaviour was found but only in girls (Li & Lee, Reference Li and Lee2010). A moderating role of 5-HTTLPR polymorphism in predicting higher depression, anxiety and somatic symptoms was also confirmed (Cicchetti et al. Reference Cicchetti, Rogosch and Sturge-Apple2007, Reference Cicchetti, Rogosch and Oshri2011), but only in interaction with sexual abuse. This relation was further moderated by MAOA activity level suggesting a three-way interaction [G × G × E].

Table 1. Summary of the studies described in this review

A-COPE, Adolescent Coping Orientation for Problem Experiences; ASSIS, Arizona Social Support Interview Schedule; AUDIT-C, Alcohol consumption alcohol use disorder identification test; CAPE, Community Assessment of Psychic Experiences; CBCL, Child Behaviour Checklist; CCQ, California Child Q-sort; CDI, Children's Depression Inventory; CTQ, Child Trauma Questionnaire; CTS, Conflict Tactics Scale; DISC, Diagnostic Interview Schedule for Children; CIDI, Composite International Diagnostic Interview; DICA-R, Diagnostic Interview for Children and Adolescents-Revised; K-SADS-PL, Schedule for Affective Disorders and Schizophrenia for School Aged Children; MCS, Maltreatment Classification System; MFQ, Mood and Feelings Questionnaire; MMCI, Maternal Maltreatment Classification Interview; MPQ, Multidimensional Personality Questionnaire; SPQ-B, Schizotypy Personality Questionnaire-Brief; STAI-T, State–Trait Anxiety Inventory; TDA, trait descriptive adjectives; TRF, Teacher Report Form; YSR, Youth Self Report.

An effect of the interaction between CRHR1 and maltreatment and early abuse on diurnal cortisol regulation was reported by Cicchetti et al. (Reference Cicchetti, Rogosch and Oshri2011): CRHR1 variations were related to cortisol dysregulation only among maltreated children, thus suggesting an allostatic load perspective on the effects of the chronic stress associated with child maltreatment on cortisol regulation and internalizing symptomatology as moderated by genetic variation.

The role of the hypothalamic–pituitary–adrenal (HPA) axis, the major biological system for stress response, was also investigated by De Young et al. (Reference De Young, Cicchetti and Rogosch2011) (Table 1), which demonstrated that CRHR1 haplotype moderated the effects of maltreatment on neuroticism, during childhood. The effect of this haplotype on neuroticism was dependent not only on the presence of maltreatment but also on the most severe type of maltreatment and number of types of maltreatment.

The BDNF was found to significantly predict youth depression in a three-way interaction with 5-HTTLPR and maltreatment history (Kaufman et al. Reference Kaufman, Yang, Douglas-Palumberi, Grasso, Lipschitz, Houshyar, Krystal and Gelernter2006), and to significantly moderate the association between childhood abuse and positive dimension of psychotic-like experiences (Alemany et al. Reference Alemany, Arias, Aguilera, Villa, Moya, Ibanez, Vossen, Gasto, Ortet and Fananas2011).

In conclusion, reported researches suggest that child maltreatment (especially physical and sexual abuse and extensive maltreatment) is a useful environmental candidate to investigate the effect of G × E interaction on the development of both developmental psychopathology and resilience. Nevertheless, considering the high variability of reported results we point out the need for future research to give greater attention to measurement and operationalization of child maltreatment, including frequency and duration, period of occurrence (developmental stage and distance from outcome) and, particularly, association with other risk or protective factors such as social context or gender. Future research should also broaden the emerging evidence that child maltreatment in interaction with specific genetic haplotypes induce significant biological changes in children (biological embedding), modifying the maturation and the operating balance of allostatic system, i.e. the biological systems responsible for maintaining physiological stability through environmental changes (Cicchetti et al. Reference Cicchetti, Rogosch and Oshri2011).