OBJECTIVE MEASURES OF EMPATHY: PROSOCIAL LEARNING AND MOOD STATES

Empathy, which is crucial for maintaining social functioning in humans, is frequently impaired in people with mental disorders. Given its fundamental role in social functioning, enhancing empathy might offer a promising way to improve treatment outcomes in individuals with mental disorders, such as major depressive disorder or suicide (Szanto et al., 2012). However, the relationship between empathy and mental disorders is not straightforward, likely due to heterogeneous definitions of empathy and the lack of objective measures. Studying a measurable component of empathy, such as prosocial learning— modifying behaviors by incorporating cues from interpersonal experiences to enhance the quality of social interactions—could help address this challenge. Using objective measures, including a computerized task to quantify degrees of empathy, this pilot study examined whether reported mood states (such as happiness, feeling challenged, and stress) were associated with learning rates for oneself and others. Specifically, the study hypothesized that positive ongoing mood (reported happiness and feeling challenged) would be positively related to learning rates for others, but not for oneself. This pilot study included seven healthy volunteers (M_age=28,SD_age=4.69; 4F/3M). Learning hyperparameters were measured using a prosocial learning task (Lockwood et al., 2016) optimized to quantify degrees of reward-seeking patterns in different social settings involving playing for the self versus others. Empathy was assessed using a Pavlovian reinforcement learning model to evaluate learning parameters, such as learning rates. Participants were instructed to perform their best during the task and report their current mood in every four trials, specifically how much they felt happiness, challenge, and stress. To determine whether learning rates for the self and others differed, we conducted a paired Wilcoxon’s test. We then explored fluctuations in mood ratings over time using Levene’s Test. To estimate the relationship between the mean and variance of mood ratings and learning rates for the self and others, we conducted a Kendall’s tau test. The results indicated that learning rates for the self and others were not significantly different (V=21, p=.297), suggesting that participants did not differ in how quickly they adapted their behavior for the task depending on whom they were playing for. Reported challenge (F(7,159)=6.23, p < .001) and stress (F(7,159)=10.63, p < .001) fluctuated over time, whereas happiness did not (F(7,159)=1.22, p=.296). Furthermore, the mean reported happiness was positively correlated with learning rates for others (r=0.90, p=.005), but not for oneself (r=0.05, p=.919). The increased learning rate for others may indicate that individuals with higher reported happiness are more attentive and motivated when playing for others, potentially serving as a marker for empathy. Contrary to our hypothesis, the mean and variance of other reported mood scores were not significantly related to learning rates for the self or others (ps > .05). These findings may indicate that reported happiness during the task is associated with empathy or that empathy may be a prerequisite for empathic behaviors, potentially serving as a cross-diagnostic marker for mental disorders. In the future study, this task will be integrated into future clinical trials to investigate whether changes in learning parameters, as a quasi-index of empathy, in patients with depression and suicide risk are associated with neural markers and ketamine treatment.

References

Lockwood, P. L., Apps, M. A., Roiser, J. P., et al. Neurocomputational mechanisms of prosocial learning and links to empathy. Proc Natl Acad Sci U S A, 2016; 113(35), 9763-9768. Szanto, K., Dombrovski, A. Y., Sahakian, B. J., et al. Social emotion recognition, social functioning, and attempted suicide in late-life depression. The American Journal of Geriatric Psychiatry, 2012; 20(3), 257-265.