To Trust Or Not To Trust: A Thermal Imaging Study On The Effects Of Social Exclusion On Decision Making

Abstract: Regulating emotions elicited by others is pivotal for social interaction. We investigated behavioural and physiological consequences of social exclusion and found that sympathetic activation triggered by ostracizing interactions is essential to learn whether to trust others and to adjust future behaviour consequently. This mechanism seems deficient in low-regulators. Supporting Summary: While emotion regulation seems to be well understood in the individual domain, information about the inherent link between emotion regulation and social contexts is meager. Here, we investigated the behavioural and physiological consequences of social exclusion in control participants (N=17) and in patients (N=16) with psoriasis, a skin disease known to be often accompanied with emotion regulation deficits. Firstly, participants faced the social induction phase by playing the Cyberball Game in which they were excluded or included by other players. Then participants played the Trust Game (TG) in the role of investor. All participants took part in both exclusion and inclusion social inductions and played the TG twice. In the TG, the investor has to decide how much of €10 to invest on familiar and unfamiliar players. The familiar (bad vs good) players were the same players previously encountered in the Cyberball Game. Participants' face temperature (peri-orbital region) during the Cyberball Game was measured by means of functional infrared thermal imaging (fITI). Face temperature in this area is known to reflect the activation of the sympathetic system. We tested whether the social inclusion vs. exclusion affected participants’ trust toward other players by entering mean investments into a Repeated Measures 2 (social induction: exclusion, inclusion) x 2 (player: familiar, unfamiliar) mixed ANOVA. Group (patient, control) was entered as a between-subjects factor and peri-orbital temperature index (i.e., temperature during social exclusion minus temperature during social inclusion) as covariate in the model. We found a statistically significant main effect of social induction, qualified by a significant interaction with peri-orbital temperature index (F(1,30) = 5.00, p = .03, partial eta-squared = .32) and also a social induction x player two-way interaction (F(1,30) = 5.39, p = .03, partial eta-squared = .15). A post hoc Bonferroni test showed that unfamiliar players were trusted significantly more after the social exclusion phase (during which participants were excluded from familiar players) respect to the inclusion one (p = .04) (during which participants were included by familiar players). A linear regression was calculated to predict a behavioural index (i.e., the differential investment on unfamiliar players taking place during social exclusion vs inclusion) based on peri-orbital temperature index. A significant regression equation was found (F(1,30) = 7.40, p = .01, R2 = .20), suggesting that trust for unfamiliar players is driven by enhanced activation of the sympathetic system during social exclusion vs inclusion phase. Interestingly, a correlation analysis showed that the behavioural adjustments triggered by peri-orbital temperature increase during social exclusion were present only in the control group (r = .54, p = .03), while patients’ decisions were not influenced by temperature change (r = .24, p = .37). This result suggests that sympathetic activation during painful social interactions is pivotal in order to learn whether to trust others and to adjust future behaviour consequently. This mechanism seems deficient in patients with emotion regulation deficits