Abstract and Introduction
Abstract
The tendency to trust and to cooperate increases from adolescence to adulthood. This social development has been associated with improved mentalizing and age-related changes in brain function. Thus far, there is limited imaging data investigating these associations. We used two trust games with a trustworthy and an unfair partner to explore the brain mechanisms underlying trust and cooperation in subjects ranging from adolescence to mid-adulthood. Increasing age was associated with higher trust at the onset of social interactions, increased levels of trust during interactions with a trustworthy partner and a stronger decline in trust during interactions with an unfair partner. Our findings demonstrate a behavioural shift towards higher trust and an age-related increase in the sensitivity to others' negative social signals. Increased brain activation in mentalizing regions, i.e. temporo-parietal junction, posterior cingulate and precuneus, supported the behavioural change. Additionally, age was associated with reduced activation in the reward-related orbitofrontal cortex and caudate nucleus during interactions with a trustworthy partner, possibly reflecting stronger expectations of trustworthiness. During unfair interactions, age-related increases in anterior cingulate activation, an area implicated in conflict monitoring, may mirror the necessity to inhibit pro-social tendencies in the face of the partner's actual levels of cooperation.
Introduction
Although humans are social by nature, the cognitive abilities that form the basis for successful social interactions are not fully developed at birth, but evolve gradually over time (Hughes, 2004; Ensink and Mayes, 2010). During the transition from adolescence to adulthood, social behaviour becomes increasingly oriented towards others (Steinberg and Sheffield Morris, 2001; Eisenberg et al., 2005; Dumontheil et al., 2010). Improved mentalizing, the sensitivity to the perspective of others, has been suggested to drive increases in trust and cooperation (King-Casas et al., 2005; Sutter and Kocher, 2007; Van den Bos et al., 2010, 2011b). The changes in social cognition and behaviour occur in parallel with structural and functional maturation of the brain. Several studies have investigated the neural correlates of social interactions, but research has only just begun to investigate the brain–behaviour association from a developmental perspective (Van den Bos et al., 2011b).
Over the last decade, combining approaches from neuroscience and economic research has lead to an interest in the neural mechanisms underlying trust and cooperation in adults (King-Casas et al., 2005, 2008; Krueger, 2008). Social cognition in action has been investigated with economic exchange paradigms, such as the trust game (Berg et al., 1995). The trust game requires mentalizing in order to infer intentions from behavioural cues of the other player and to appreciate how own behaviour (e.g. lowering trust in response to trustworthiness) leads to a reputation with others. During the trust game, the first player (investor) receives an amount of money from the experimenter and can choose to cooperate (i.e. share any part of the money) with the second player (trustee) or to defect (i.e. keep the money). The shared amount is multiplied and the trustee decides whether to return any part of this amount or whether to keep the money. The best pay-offs for both players occur when they cooperate. However, the trustee yields the highest pay-off by defecting. Thus, to share money, the investor needs to trust in the good intentions of the trustee. Despite different predictions from classic economic theory, investors typically send a share of 50% or more of their initial endowment; this signal of trust is generally reciprocated by the trustee (Gintis, 2000; Camerer, 2003; Krueger, 2008). In multi-round versions of the trust game, the degree of trustee reciprocity is a strong predictor for subsequent decreases or increases in investor trust (King-Casas et al., 2005), showing that trust or the expectation about future behaviour is modulated by the actual behaviour of the game partner.
Previous functional magnetic resonance imaging (fMRI) research with the trust game demonstrated activation in brain regions important for mentalizing, reward learning, cognitive control and emotional processing (Rilling et al., 2004; Delgado et al., 2005; King-Casas et al., 2005; Krueger et al., 2007; Krueger, 2008; Van den Bos et al., 2011b). It has been postulated that the reward network, extending from the striatum and specific frontal regions, is involved in the motivation to cooperate and that brain networks of cognitive control and social cognition modulate this motivation in response to contextual information (Declerck et al., 2011). Cognitive control is important for the adaptation of behavioural patterns in response to new evidence, e.g. behavioural feedback from interaction partners (Kerns, 2004; Magno et al., 2006). The social cognitive network supports mentalizing, the process of interpreting others' social signals and is important to minimize betrayal (Van Overwalle, 2009; Declerck et al., 2011).
Although cooperation seems to be the preference of adults, developmental studies using trust games suggest a tendency to invest lower amounts and to have less reciprocal interactions in adolescents (Sutter and Kocher, 2007; Van den Bos et al., 2011a, b). This changing quality of social interactions has been attributed to the lower propensity of adolescents to mentalize and a subsequently reduced sensitivity to others' social signals (Dumontheil et al., 2010; Van den Bos et al., 2011b). Studies have begun to elucidate how these differences in social behaviour and cognition are reflected in differential activation of networks subserving social interactions (Van den Bos et al., 2011b). In the study by Van den Bos et al. (2011b), a sample of 62 participants between 12 and 22 years took the trustee role in a two-choice trust game with a generally trustworthy investor. Within this sample, age was unrelated to the degree of reciprocity towards the investor. Yet, indicative of increased mentaliszing, the sensitivity to the degree of risk that the investor took during decision making significantly increased with age. The neuroimaging results showed age-related increases in brain activation in the left temporo-parietal junction (TPJ) and the right dorsolateral prefrontal cortex in response to investor trust. The TPJ has been suggested to play a role in identifying goals and intentions behind others' behaviour (Mitchell et al., 2005; Van Overwalle, 2009). Previous research showed age to be associated with higher TPJ activity during simple mentalizing tasks (Blakemore et al., 2007) and self-referential processing (Pfeifer et al., 2007). The involvement of the dorsolateral prefrontal cortex was hypothesized to reflect the age-related regulation of selfish responding. Age was also associated with decreases in activation in the anterior medial prefrontal cortex during defection, when compared with reciprocation of investor trust. This area has also been hypothesized to underlie mentalizing processes (Krueger, 2008), specifically those that include thinking about how oneself is perceived by others (Amodio and Frith, 2006; Frith and Frith, 2008). However, previous developmental research also found this area more active during self-related processing when compared with social processing (Blakemore et al., 2007; Pfeifer et al., 2007).
The previously described study presents initial behavioural evidence in support of age-related increases in the sensitivity to others' perspectives, as indicated by an increased degree of reciprocity when the interaction partner made high-risk investment decisions. However, it remains unclear whether the sensitivity to the other person's behavioural cues also increases with age and whether this underlies the age-related development towards a behavioural default of trust and cooperation. To elucidate this question, this study investigated age-related changes in investor behaviour as a function of partner reciprocity. Participants played two multi-round trust games with anonymous hypothetical game partners, one with a cooperative and one with an unfair decision-making style. If sensitivity to the other person's behavioural cues increases with age, more pronounced increases or decreases in the levels of trust with age should occur in response to cooperation or unfair behaviour by the trustee, respectively.
There is limited research examining social cognitive and behavioural changes during social interactions across broader age ranges. However, previous research indicates that social (cognitive) processes continue to change into adulthood (Dumontheil et al., 2010). Therefore, this study included a sample of participants ranging from adolescence to mid-adulthood. We hypothesized that age would be associated with higher trust and an increased sensitivity to social signals of others and that this would be reflected in (i) higher initial investments and (ii) higher investments throughout interactions with a cooperative, but (iii) lower investments towards an unfair game partner, as any increased sensitivity to the behavioural cues of the other person will likely involve better mentalizing skills and/or better social reward learning. We expected at the neural level (i) age-related increases in brain activation within frontal and temporo-parietal brain areas implicated in mentalizing (e.g. medial prefrontal cortex, TPJ and precuneus) and (ii) age-related decreases in activation in social reward-related areas (e.g. ventromedial prefrontal and orbitofrontal cortex and caudate nucleus) during interactions with a cooperative trustee as a consequence of age-related increases in expectations of the trustworthiness of others. During interactions with an unfair trustee, we similarly expected increases in brain areas implicated in mentalizing, but in addition (iii) age-related increases in activation in cognitive control regions (e.g. anterior cingulate cortex and dorsolateral prefrontal cortex) because of the need to suppress the default intention to invest.
Soc Cogn Affect Neurosci. 2014;9(4):395-402. © 2014 Oxford University Press
