Humans make rapid decisions of trustworthiness based on facial appearance. Williams Syndrome (WS) is a genetic disorder associated with hypersociability toward strangers, suggesting a disruption in trust assessment. We recorded eye-movements in neurotypically-developed (TD) participants (N = 21) and in patients with WS (N = 22) as we presented pairs of computer-generated faces, pre-rated as trustworthy or untrustworthy. In a spontaneous visual preference task, TD participants gazed significantly longer at trustworthy faces while patients with WS showed no preferences for either face-type. Next, in a decision task, participants selected the face that they judged consistent with one of three social descriptions relating to high levels of trust. Both groups gazed longer at trustworthy faces and made correct matches with the corresponding social descriptions. Improvements by patients with WS in the decision task were contingent on outcomes measured on the emotional NEPSY scale. We conclude that while spontaneous representation of trustworthiness is impaired in WS, top-down mechanisms of trust recognition appear partially preserved.

人类根据面部表情快速做出可信赖度的决策。威廉姆斯综合症（WS）是一种与对陌生人的超社交能力相关的遗传性疾病，表明信任评估受到干扰。我们记录了成对的计算机生成的面部，并被预先评定为可信赖或不可信赖，并记录了神经型发育（TD）参与者（N = 21）和WS患者（N = 22）的眼球运动。在自发的视觉偏好任务中，TD参与者凝视可信赖面孔的时间明显更长，而WS患者对这两种面孔类型均无偏好。接下来，在一项决策任务中，参与者选择了自己认为与高信任度相关的三个社会描述之一相符的面孔。两组人都对可信赖的面孔凝视了更长的时间，并与相应的社会描述进行了正确的匹配。WS患者在决策任务中的改善取决于以情绪NEPSY量表衡量的结局。我们得出的结论是，虽然WS中自发性的可信度表示受到损害，但信任识别的自上而下的机制似乎保留了一部分。