less heritable (30–40%), with less stability in genetic influences across ages (Bartels et al., 2004; Burt & Neiderhiser, 2009; Eley et al., 2003). However, different types of conduct disorder may themselves have different etiologies. For example, conduct disorder, which is comorbid with ADHD, has been found to be more heritable than other forms of conduct disorder (Thapar et al., 2001).
In relation to the etiology of bullying, which can also be considered as a subset of antisocial or aggressive behavior, twin studies demonstrated that 60–70% of the variance in bullying behavior can be influenced by genetic factors, with the remainder due to nonshared environmental effects (Ball et al., 2008; Veldkamp et al., 2019). Similar high genetic influences were found for bullying victimization. These findings suggest that if siblings from one family bully other children or are bullied by other children, it is more likely to be due to shared genetic inheritance than the family environment (such as parenting style).
Adoption studies
Another family‐based approach that allows estimation of the effects of genetic and environmental influences on individual differences in a trait is the adoption approach. In adoption studies, phenotypic similarities (similarities on a specific trait) can be compared between adopted children and their biological and adoptive parents; and between siblings from biological families and siblings from adoptive families. Adoption studies are also a powerful tool to study interactions between genetic and environmental factors.
In line with twin studies, adoption studies have demonstrated substantial effects of genetic factors on the development of social behavior. Studies of externalizing problems in adopted children of age 10–15 years, and their biological and nonbiological siblings, showed that externalizing behavior is mostly explained by genetic factors (65%). Similarly to the results of twin studies, heritability of aggressive behavior (70%) was higher than heritability of delinquent behavior (39%) (Van den Oord et al., 1994). However, in the Colorado Adoption Project, which also compared biological and nonbiological siblings, heritability of aggression (24–49%) and delinquent behavior (17–36%), assessed by parental and teachers’ reports, was lower (Deater‐Deckard & Plomin, 1999). Another example of an adoption study is the Early Growth and Development Study (EGDS), a US nationwide, prospective study of birth parents and adoptive families (https://egds.la.psu.edu/). A study that used EGDS data from 346 linked triads (birth mother, adoptive parents, adopted child) in order to assess externalizing behavioral problems, such as aggression and defiance, when children were 18, 27, and 54 months of age (Kerr et al., 2013). The results of this study indicated that biological mothers’ histories of major depression and adult antisocial behavior were found to confer risk for child externalizing behaviors at age 18 months, consistent with the role of genetic influences on child behavioral problems.
With trends toward selective adoption and the diminishing frequency of adoptions in many developed countries, including the United Kingdom and the United States, adoption studies are becoming a less feasible method of genetically informative investigations. However, the increased rate of reconstituted families (families composed of both siblings and half siblings) may offer a new way to evaluate the role of genetic factors in the transmission of complex disorders (Risch et al., 2014).
To summarize, there is growing evidence, from both molecular genetic studies and family‐based (twin and adoption) studies, for the complex polygenic architecture of sociobehavioral phenotypes, such as antisocial behavior. Future studies are needed to uncover the role of specific genetic variants across the whole genome, and their complex interplay with various environmental factors.
The Role of Environments Factors in Social Development
In behavioral genetics, the environment is viewed as any influences on a specific phenotype (trait) that are not genetic (not related to DNA) and that contribute to individual differences in this phenotype in a population. Therefore, this definition of the environment is broader than the one that is usually used in psychology. For example, this definition may include influences in prenatal and postnatal periods, such as nutrition, illnesses, hormonal levels, and social relationships with family and peers.
One of the goals of behavioral genetic research is to identify specific environmental factors that contribute to individual differences in psychological traits, including social development (Reiss et al., 2000; Turkheimer & Waldron, 2000). In order to do this, genetically informative designs, such as twin and adoption designs, have been employed (Plomin & Daniels, 1987; Reiss et al., 2000; Rutter, 2006; Rutter et al., 2001; Turkheimer & Waldron, 2000). Many genetically informative studies use multiple informants (children, parents and other family members, peers, teachers, researchers) to measure environmental factors, such as parent–child relationships, sibling relationships, and relationships with peers and teachers.
The twin method allows for estimation of the relative contribution of two types of environment: shared (common) environment that contributes to the similarity between the two twins in a pair on a particular trait; and nonshared (unique) environment that contributes to the dissimilarity between the two twins. As twins grow up together they objectively share many aspects of their environment, including parents, socioeconomic conditions and often schools, classrooms, teachers, and peers (Plomin et al., 2013). The question of why children in the same family are very different from each other has been extensively researched in behavioral genetics (Plomin, 2011; Plomin et al., 2001; Plomin & Daniels, 1987). One of the findings is that objectively shared factors more often than not end up as nonshared effects. For example, socioeconomic status (SES) of the family, objectively shared by the two twins, may actually lead to the differences between the twins if they perceive or respond to this status differently (Plomin et al., 2001). A low SES may motivate one child to strive for achievement, but lead to lower motivation and achievement in another child. Such subjective reactions may form under the influence of many factors, including genetics (see the section “The Role of Gene–environment Interplay in Social Development”).
The results of a meta‐analysis of 43 studies demonstrated that nonshared environment accounts for ~50% of variability of many behavioral traits. However, none of the measured nonshared environmental factors (“objective” environments) has explained more than 2% of phenotypic variance (Turkheimer & Waldron, 2000). Since the publication of this meta‐analysis, multiple studies have been conducted searching for objective nonshared environmental factors involved in behavior of children and adolescents in different contexts (Asbury et al., 2003; Beaver, 2008; Deater‐Deckard et al., 2001). For example, a sample of 289 monozygotic twin pairs from the National Longitudinal Study of Adolescent Health (Add Health) in the United States was used to examine the effects of the nonshared familial environment on delinquent involvement, adult criminal behavior, levels of self‐control, and contact with antisocial peers (Beaver, 2008). Analysis revealed that one nonshared family environment – maternal disengagement – was associated with delinquent involvement and the development of self‐control. The remaining nonshared familial environments were not associated with social outcomes.
An extension of the twin design is the children‐of‐twins (COT) design that is based on samples of monozygotic (MZ) and dizygotic (DZ) twins with children. The offspring of MZ twins are as related to their parent’s co‐twin as they are to their own parent (they share 50% of their DNA). In contrast, the offspring of DZ twins share 25% of their genetic variance with their parent’s co‐twin. By comparing MZ avuncular correlations (correlations between uncle/aunt and niece/nephew) with DZ avuncular correlations, it is possible to estimate the role of genetic factors in explaining intergenerational associations. It is also possible to estimate the extent to which parent–child associations remain after accounting for genetic transmission. D'Onofrio et al. (2007), using this COT method, found that the effect of parental divorce on offspring substance use problems was significant even when genetic and other environmental risks were controlled for.
A large‐scale longitudinal study, the NEAD
