Oxytocin enhances pupil dilation and sensitivity to 'hidden' emotional expressions

Abstract

Sensing others' emotions through subtle facial expressions is a highly important social skill. We investigated the effects of intranasal oxytocin treatment on the evaluation of explicit and 'hidden' emotional expressions and related the results to individual differences in sensitivity to others' subtle expressions of anger and happiness. Forty healthy volunteers participated in this double-blind, placebo-controlled crossover study, which shows that a single dose of intranasal oxytocin (40 IU) enhanced or 'sharpened' evaluative processing of others' positive and negative facial expression for both explicit and hidden emotional information. Our results point to mechanisms that could underpin oxytocin's prosocial effects in humans. Importantly, individual differences in baseline emotional sensitivity predicted oxytocin's effects on the ability to sense differences between faces with hidden emotional information. Participants with low emotional sensitivity showed greater oxytocin-induced improvement. These participants also showed larger task-related pupil dilation, suggesting that they also allocated the most attentional resources to the task. Overall, oxytocin treatment enhanced stimulus-induced pupil dilation, consistent with oxytocin enhancement of attention towards socially relevant stimuli. Since pupil dilation can be associated with increased attractiveness and approach behaviour, this effect could also represent a mechanism by which oxytocin increases human affiliation.

Keywords: emotion; empathy; hormones; locus coeruleus; pupillometry; social.

Figures

Fig. 1

Overview of study design. After administration of nasal spray containing oxytocin or placebo, participants underwent a test protocol consisting of face stimuli presented simultaneously with either soft stroking touch or vibration. The visual stimuli included five expression types: explicit anger, implicit (hybrid) anger, neutral, implicit (hybrid) happiness and explicit happiness. Implicit expressions contained high-frequency visual information from a neutral expression and low-frequency visual information from the same person expressing either anger or happiness. These stimuli are perceived as neutral, but were previously shown to evoke a core impression such that faces containing implicit happiness are perceived as more friendly than faces containing implicit anger, fear or sadness (Laeng et al., 2010). Participants rated perceived emotion, social characteristics or tactile characteristics after each stimulus pair. Images were adapted from the Karolinska Directed Emotional Faces—KDEF (CD-ROM), by D. Lundqvist, A. Flykt, & A. Ohman, 1998, Stockholm, Sweden: Department of Clinical Neuroscience, Psychology section, Karolinska Institutet. Reprinted with permission.

Fig. 2

(A) Oxytocin treatment caused a stimulus-congruent ‘sharpening’ of perceived anger and happiness, such that angry expressions were perceived as more angry and less happy, whereas happy expressions were perceived as more happy and less angry. Oxytocin-induced ‘sharpening’ (i.e. oxytocin-induced changes in perceived mood, calculated as the between-session difference in mean ratings for each participant and facial expression) is represented on the y axis. (B) This ‘sharpening’ effect of intranasal oxytocin treatment was evident across explicitly and implicitly presented expressions, as illustrated here with a depiction of the raw scores. Error bars represent the standard error of the mean. **P < 0.01, *P < 0.05.

Fig. 3

(A and B) A median-split analysis based on baseline emotional sensitivity scores illustrates the relationship between sensitivity to subtle differences in emotional expressions and oxytocin enhancement of performance on this task. A high score on this measure indicates high sensitivity to differences between the two ‘hybrid’ images containing hidden anger or happiness in terms of perceived anger and perceived happiness. Participants with a high emotional sensitivity score performed only marginally better in this task after oxytocin treatment (Ps > 0.73). In contrast, oxytocin significantly improved the task performance of those who did not reliably report more anger for the implicitly angry faces (relative to implicitly happy faces, chart A) or more happiness for the implicitly happy expressions (relative to implicitly angry expressions, chart B) in the placebo condition (both Ps < 0.01). (C) A similar relationship was found when task-induced pupil dilation was used as an independent moderator, such that those with greater task-induced pupil dilation showed the largest improvement in emotional sensitivity after oxytocin treatment. Error bars represent the standard error of the mean.

Fig. 4

Oxytocin pre-treatment caused significantly larger stimulus-induced pupil dilation (A). Pupil responses in the oxytocin session were significantly larger than in the placebo session during the 1000–3000 ms interval after stimulus onset. Task-induced pupil dilation correlated with individual variability in emotional sensitivity towards differences in subtle expressions (r = –0.376, P < 0.01). Those with the lowest sensitivity showed larger pupil dilation during stimulus presentation, consistent with higher task difficulty for these participants. As illustrated by the median split analysis (C), oxytocin’s beneficial effects on emotional sensitivity in the low sensitivity group were not underpinned by large increases in pupil dilation. In contrast, we found a trend towards a group-by-treatment interaction driven by a larger oxytocin-induced increase in pupil dilation in the high emotional sensitivity group.