Since behavioral genetic research established the influence of both environmental and genetic influences on a spectrum of traits and behaviors (Plomin et al., Reference Plomin, DeFries, McClearn and McGuffin2008), the age-old question of nature versus nurture has been, to a great extent, replaced or extended by questions regarding the specific genetic and environmental effects involved in each trait\behavior. This advancement opened the door to research on the nature of gene-environment interactions (GxE; e.g., risk vs. differential susceptibility, and possibility of resilience to environmental influences; Caspi et al., Reference Caspi, Sugden, Moffitt, Taylor, Craig, Harrington, McClay, Mill, Martin, Braithwaite and Poulton2003; Ellis et al., Reference Ellis, Boyce, Belsky, Bakermans-Kranenburg and van Ijzendoorn2011), and on potential epigenetic mediators of these interactions (McGowan et al., Reference McGowan, Sasaki, D'Alessio, Dymov, Labonté, Szyf, Turecki and Meaney2009). Similarly, more attention has been drawn to gene-environment correlations (rGE; Plomin et al., Reference Plomin, DeFries and Loehlin1977) as an important interplay of genetic and environmental influences (Jaffee & Price, Reference Jaffee and Price2007). These progressively more complex questions require a broader coverage of phenotypes, genes, and environments, nested in a longitudinal design, in order to understand development. With these questions in mind the Longitudinal Israeli Study of Twins (LIST) was designed (Knafo, Reference Knafo2006).
The LIST aims to advance knowledge on the development of individual differences in children's social behavior, using a comprehensive cognitive-developmental-social-genetic design. The LIST is structured to follow children and their parents every two years or less from the age of 3 years, and to collect four types of data: questionnaire, observational, physiological (added at the age of 7 years), and genetic (some epigenetic data are also available on a sub-sample). The scope of the data enables an investigation of rGE, GxE, and the crossroads of genes and environment (epigenetics).
Data Collection
The Ministry of the Interior has provided data on families of monozygotic (MZ) and dizygotic (DZ) twins. When the twins reached the age of 3 years, questionnaires were sent to the families, which included questions on the pregnancy, twins’ behavior—particularly pro-social behavior—twins’ characteristics, the relationship between the twins, demographic details, socioeconomic status, and questions regarding the twins’ zygosity. At ages 3.5, 5, 6, and 7 years, families living in Jerusalem and nearby surroundings were invited to partake in an experimental session at the laboratory (e.g., Avinum, Ebstein, & Knafo, Reference Avinun, Ebstein and Knafo2012; Pener-Tessler et al., Reference Pener-Tessler, Avinun, Uzefovsky, Edelman, Ebstein and Knafo2013 Kavé, Shalmon, & Knafo, in press). At the age of 6 years, we were also able to cover additional areas in Israel and conduct the experiment at families’ homes when necessary.
Table 1 depicts sample sizes for questionnaire and observational data separately according to age and zygosity. Thus far data have been collected from 1,636 twin families, 19.7% MZ twins, 71.1% DZ twins, and 9.2% of unknown zygosity. The high proportion of DZ twins is due to the relatively high number of Israeli families who use assisted reproductive technologies. DNA data collection is ongoing and at the time of this report is available on 37.7% of the sample. Zygosity is assigned based on parental questionnaire data (Goldsmith, Reference Goldsmith1991) or 10 genetic polymorphisms when DNA data are available. As found in previous research (Price et al., Reference Price, Freeman, Craig, Petrill, Ebersole and Plomin2000), the algorithm used to assign zygosity with a questionnaire method was in 95% agreement with DNA results.
TABLE 1 LIST MZ and DZ Pairs and Data Collected According to Age and Data Type
Participant numbers do not include pairs whose zygosity has not been verified or estimated yet. Data collection is ongoing.
Behavioral measurements
Table 1 also describes the types of data that have been collected (for further details see Knafo, Reference Knafo2006). As can be seen, data on pro-social behavior and empathy, two key variables in LIST, have been collected at each time point using both questionnaire and observational methods. Other measured attributes and characteristics, in addition to their stand-alone importance, have been linked to children's social behavior and were therefore included in LIST [e.g. parenting (see Knafo & Plomin, Reference Knafo and Plomin2006a); cognitive ability (see Zahn-Waxler et al., Reference Zahn-Waxler, Iannotti, Chapman, Rubin and Ross1982); and temperament (see Russell et al., Reference Russell, Hart, Robinson and Olsen2003)].
Genetic data
Since the focus of LIST is social, and particularly pro-social behavior, there are three neurobiological systems of special interest: dopaminergic, serotonergic, and oxytonergic\vasopressinergic. Dopaminergic brain areas are associated with reward (Beninger & Miller, Reference Beninger and Miller1998; Girault & Greengard, Reference Girault and Greengard2004; Schultz, Reference Schultz2010), and have been shown to be activated by decisions to donate to charities, thus suggesting their role in the ‘warm glow’ feeling that promotes other-oriented\pro-social behaviors (Moll et al., Reference Moll, Krueger, Zahn, Pardini, de Oliveira-Souza and Grafman2006). Serotonin has been associated with various characteristics related to social behavior such as decision-making (Stoltenberg & Vandever, Reference Stoltenberg and Vandever2010), aggression (Kuepper et al., Reference Kuepper, Alexander, Osinsky, Mueller, Schmitz, Netter and Hennig2010), and neuroticism (Munafò et al., Reference Munafò, Clark, Roberts and Johnstone2006). The oxytocin and vasopressin nonapeptides have also been associated with social behavior in various species (Donaldson & Young, Reference Donaldson and Young2008; Ebstein et al., Reference Ebstein, Israel, Chew, Zhong and Knafo2010). Thus, main candidate genes for our molecular genetic studies include the dopamine D4 receptor gene (DRD4), the dopamine D5 receptor gene, the serotonin transporter gene, the catechol-O-methyltransferase gene, the arginine vasopressin receptor 1A gene (AVPR1A), and the oxytocin receptor gene.
Recent Major Achievements in LIST
The first phases of LIST (ages 3–5 years) focused on empathy and on pro-social behavior, defined as a voluntary act aimed at benefiting others (Eisenberg et al., Reference Eisenberg, Fabes, Spinrad and Eisenberg2006). The three main forms of pro-sociality, namely sharing, caring, and helping, can all be initiated in response to internal or external motivations (self-initiated or compliant pro-social behavior, respectively). Self-initiated pro-social behavior can be seen as more advanced in terms of cognitive development as it requires going one step further than conforming to schematic or social demands (Knafo et al., Reference Knafo, Steinberg and Goldner2011b).
In accordance with previous research (Eisenberg et al., Reference Eisenberg, Cameron, Tryon, Staub, Bar-Tal, Karylowski and Reykowski1984) our results from age 3.5 years have shown that self-initiated and compliant pro-social behavior do not inter-correlate (Knafo et al., Reference Knafo, Israel and Ebstein2011a), implying these behaviors represent inherently different constructs. In the same study, it was also shown that genetic influences account for a substantial proportion of variance in observed self-initiated (43%) and compliant pro-social behavior (34%) (Knafo et al., Reference Knafo, Israel and Ebstein2011a). The rest of the variance was accounted for by non-shared environment. We found a similar pattern with observed empathy (Knafo et al., Reference Knafo, Zahn-Waxler, Davidov, Van Hulle, Robinson and Rhee2009). These results join the young but growing literature showing that genetic factors contribute to children's and adolescents’ pro-sociality (Gregory et al., Reference Gregory, Light-Häusermann, Rijsdijk and Eley2009; Hur & Rushton, Reference Hur and Rushton2007; Knafo & Israel, Reference Knafo, Israel, Mikulmcer and Shaver2009; Knafo & Plomin, Reference Knafo and Plomin2006b; Knafo et al., Reference Knafo, Zahn-Waxler, Davidov, Van Hulle, Robinson and Rhee2009; Scourfield et al., Reference Scourfield, John, Martin and McGuffin2004; Volbrecht et al., Reference Volbrecht, Lemery-Chalfant, Aksan, Zahn-Waxler and Goldsmith2007; Zahn-Waxler et al., Reference Zahn-Waxler, Schiro, Robinson, Emde, Schmitz, Emde and Hewitt2001; but also see van IJzendoorn et al., Reference van IJzendoorn, Bakermans-Kranenburg, Pannebakker and Out2010).
In LIST, the influence of maternal behavior was examined on compliant and self-initiated pro-social behavior (Knafo et al., Reference Knafo, Steinberg and Goldner2011a). Maternal positivity, negativity, and unreasoned punishment did not correlate significantly with preschoolers’ pro-social behavior. However, a gene-environment interaction was found with variation in the exon III repeat region of the DRD4 gene. Only the pro-social behavior of children, who were carriers of the 7-repeat allele, was correlated with mothers’ self-reported parenting (Knafo et al., Reference Knafo, Steinberg and Goldner2011a). A similar pattern was seen with children's empathic concern (Knafo & Uzefovsky, in press), a vicarious emotional response to others, which is closely linked to pro-social behavior. Maternal negativity related negatively to observed empathy toward an examiner at 3.5 years, but only among children carrying the 7-repeat allele.
Interestingly, as further shown by our group (Fortuna et al., Reference Fortuna, van Ijzendoorn, Mankuta, Kaitz, Avinun, Ebstein and Knafo2011b), the same seems to hold true when the roles are reversed, with mothers as the DRD4 7-repeat allele carriers and the children as their surrounding environment. Higher levels of child medical risk at birth (e.g., gestational age at birth, birth weight, and length of stay at the neonatal intensive care unit) were associated with less sensitive parenting only among mothers carrying the 7-repeat allele. Moreover, mothers who were carriers of the 7-repeat allele and whose children scored low on the risk index were observed to have the highest levels of sensitivity. These results support previous research (Ellis et al., Reference Ellis, Boyce, Belsky, Bakermans-Kranenburg and van Ijzendoorn2011) showing that the 7-repeat allele is a differential susceptibility allele, moderating the way the environment influences behavior, for better or for worse. Birth weight, which, as mentioned, is an index of medical risk, was also shown to be a predictor of children's conduct problems in discordant twins (Mankuta et al., Reference Mankuta, Goldner and Knafo2010).
The data collected in LIST also enabled us to be the first to associate preschoolers’ altruistic behavior with a variant of the RS3 polymorphism in AVPR1A (Avinun et al., Reference Avinun, Israel, Shalev, Gritsenko, Bornstein, Ebstein and Knafo2011), which was previously associated with autism (Kim et al., Reference Kim, Young, Gonen, Veenstra-VanderWeele, Courchesne, Courchesne, Lord, Leventhal, Cook and Insel2002), and to associate altruism in adults with variation in the oxytocin receptor gene (Israel et al., Reference Israel, Lerer, Shalev, Uzefovsky, Riebold, Laiba, Bachner-Melman, Maril, Bornstein, Knafo and Ebstein2009).
Peer relationships are also a key factor in the acquisition of social skills and traits, to the point that they were claimed to have a larger influence than parenting (Harris, Reference Harris1998). We investigated the relationships between children's peer problems and three temperamental dimensions: negative emotionality (an individual's tendency to experience distress), sociability (an individual's ability to enjoy interpersonal contexts), and activity level. Results showed that children characterized by negative emotionality and low sociability and activity level tended to have greater peer problems. Moreover, bivariate genetic analyses showed that most of the relationship between temperament and peer problems was due to overlapping genetic factors which contributed to both traits (Benish-Weisman et al., Reference Benish-Weisman, Steinberg and Knafo2010). Interestingly, a mirror image of this pattern was found with regard to pro-social behavior, which related negatively to negative emotionality and positively to activity level and sociability (Knafo & Israel, Reference Knafo, Israel, Zentner and Shiner2012).
The study of twin relationships may have important implications not only for understanding the interesting social dynamics of this growing group of children, but also for the assumptions and generalizability of behavioral genetic research, and yet it remains relatively scarce. A comparison between maternal reports of 3-year-old DZ and MZ twins’ rivalry, conflict, closeness, and dependence, has shown that MZ twins (including those mistakenly considered as DZ by their mothers) were perceived as closer and more co-twin dependent relative to DZ twins. Same-sex twins were reported to have lower dependence and higher levels of conflict and rivalry than opposite-sex dyads. In addition, as compared with a matched sample of non-twin siblings, DZ twins were described as closer and more co-dependent (Fortuna et al., Reference Fortuna, Goldner and Knafo2011a). These findings lend support to theories regarding the effects of genetic similarity and gender in shaping relationships, and speak to the special nature of twins’ relationships.
Future Directions
The longitudinal aspect of the collected data will enable us to examine stability and change throughout development in various pro-social, temperamental, and parental characteristics. The second phase of LIST (ages 6–7 years, and a planned age 8–9 wave) has two main goals. The first is to provide an in-depth investigation of gene–environment correlations, a subject that has received very little attention (Knafo & Jaffee, Reference Knafo and Jaffee2013). We will rely on the inclusion of parental behavior in the study design and the collection of DNA from parents and twins (e.g., Pener-Tessler et al., Reference Pener-Tessler, Avinun, Uzefovsky, Edelman, Ebstein and Knafo2013). A first report from LIST found evidence for evocative gene-environment correlations, indicated by heritability of observed maternal behavior, in mothers’ intrusiveness but not in their warmth (Knafo, Reference Knafo, Kourkoutas and Erkman2011).
The second main goal is to further understand pro-social development. For example, we are currently investigating the role of children's values in the development of pro-social and empathic behaviors, and their association with physiological biomarkers of the autonomic nervous system. In addition to studying potential phenotypic relationships, we will investigate the role of genetics in these relationships as well as parent-child cycles of influence.
Finally, in addition to the effects of the candidate genes mentioned above, we are interested in how gene–environment interactions manifest in epigenetic changes in the genome. The collection of genetic, epigenetic, biological, and environmental data is aimed at providing an integrated view of the developmental processes underlying social behavior.
Acknowledgments
We thank Naama Gilat, Noa Gordon Assayag, Rivka Shir, and Florina Uzefovsky for their help. The LIST was founded by Grant No. 31/06 from the Israel Science Foundation. Further support was obtained by a grant from The Science of Generosity Initiative, University of Notre Dame, funded by the Templeton Foundation, and by Starting Grant no. 240994 from the European Research Council, both to Ariel Knafo.
