Boosting Empathy and Compassion Through Mindfulness-Based and Socioemotional Dyadic Practice: Randomized Controlled Trial With App-Delivered Trainings

BACKGROUND: Contemplative trainings have been found to effectively improve social skills such as empathy and compassion. However, there is a lack of research on the efficacy of app-delivered mindfulness-based and dyadic practices in boosting socioaffective capacity. OBJECTIVE: The first aim of this study was to compare a novel app-delivered, partner-based socioemotional intervention (Affect Dyad) with mindfulness-based training to foster empathy and compassion for the self or others. The second aim of this study was to investigate the underlying mechanisms of these effects. METHODS: This randomized controlled trial included socioemotional and mindfulness-based interventions and a waitlist control group, which received socioemotional training after the postintervention assessment. We used linear mixed-effects models to test intervention effects on self-report measures and an ecologically valid computer task of empathy, compassion for the self and others, and theory of mind. Moderated mediation models were used to investigate whether changes in acceptance, empathic distress, empathic listening, interoceptive awareness, and mindfulness served as underlying psychological processes of intervention effects. RESULTS: In 218 participants (mean age 44.12, SD 11.71 years; 160/218, 73.4% female), we found all interventions to have positive effects on composite scores for compassion toward the self (β(socioemotional)=.44, P<.001; β(waitlist socioemotional)=.30, P=.002; β(mindfulness-based)=.35, P<.001) and others (β(socioemotional)=.24, P=.003; β(waitlist socioemotional)=.35, P<.001; β(mindfulness-based)=.29, P<.001). Compassion measured with the computer task did not change significantly but showed a trend toward increase only in socioemotional dyadic practice (β(socioemotional)=.08, P=.08; β(waitlist socioemotional)=.11, P=.06). Similarly, on the empathic concern subscale of the Interpersonal Reactivity Index, a nonsignificant trend toward increase was found in the socioemotional intervention group (β(socioemotional)=.17; P=.08). Empathy significantly increased in both socioemotional groups (β(socioemotional)=.16, P=.03; β(waitlist socioemotional)=.35, P<.001) and the mindfulness-based group (β(mindfulness-based)=.15; P=.04). The measures of theory of mind did not change over time. In the mindfulness-based group, the increase in self-compassion was mediated by a decrease in empathic distress (indirect effect ab(mindfulness-based)=0.07, 95% CI 0.02-0.14). In the socioemotional group, an increase in self-compassion could be predicted by an increase in acceptance (β(socioemotional)=6.63, 95% CI 0.52-12.38). CONCLUSIONS: Using a multimethod approach, this study shows that app-delivered socioemotional and mindfulness-based trainings are effective in fostering compassion for the self and others in self-report. Both low-dose trainings could boost behavioral empathy markers; however, the effects on behavioral and dispositional markers of compassion only trended after dyadic practice, yet these effects did not reach statistical significance. Training-related increases in self-compassion rely on differential psychological processes, that is, on improved empathic distress regulation through mindfulness-based training and the activation of a human care– and acceptance-based system through socioemotional dyadic training. TRIAL REGISTRATION: ClinicalTrials.gov NCT04889508; https://clinicaltrials.gov/ct2/show/NCT04889508