Abstract:

A considerable number of studies have discussed the association between childhood maltreatment and empathy, but the results have been mixed. Theoretically, there are four main arguments regarding the association between childhood maltreatment and empathy. Attachment theory suggests that childhood maltreatment predisposes individuals to an insecure attachment style that is detrimental to empathy development. The facial feedback hypothesis and “like-me” hypothesis suggest that neglected children have fewer opportunities to imitate others’ facial expressions, and show deficits in empathy. However, the perception-action model argues that individuals with childhood maltreatment are more likely to empathize with others who have traumatic experiences, while the Russian doll model implies that the association between childhood maltreatment and empathy may not be linear. Empirically, current research has reported an inconsistent correlation between childhood maltreatment and empathy, with r values ranging from −0.451 to 0.86. Therefore, the present meta-analysis aimed to estimate the extent to which childhood maltreatment is associated with empathy and whether these associations vary depending on the study or sample characteristics such as the type of childhood maltreatment, gender, and age.

A systematic literature review was conducted using Web of Science, ScienceDirect, PubMed, PsycARTICLES, CNKI (China National Knowledge Infrastructure), CSTJD (China Science and Technology Journal Database) and WFD (Wan Fang Data). The literature screening process is shown in Figure 1. Three-level meta-analyses were performed using R to synthesize the effect sizes and conduct moderator analyses. Publication bias was assessed using funnel plots and Egger's regression tests. No significant publication bias was observed in any of the studies.

A total of 46 studies (N = 23039 participants) producing 352 effect sizes were included. The results showed that the correlation between childhood maltreatment and empathy was significantly negative but only to a small extent (r = −0.076, 95% CI [−0.117, −0.035]). Table 1 showed the results of moderator analysis, which revealed that the association between childhood maltreatment and empathy was moderated by the type of childhood maltreatment. The association between childhood maltreatment and empathy was stronger for physical neglect (r = −0.095) and emotional neglect (r = −0.128) than for physical abuse (r = 0.005). Furthermore, the association between childhood maltreatment and empathy was moderated by the dimensions of empathy. Specifically, the association between childhood maltreatment and empathy was negative for perspective-taking (r = −0.127), fantasy (r = −0.044), and empathic concern (r = −0.148), but positive for personal distress (r = 0.153). In addition, the mean age of the participants moderated the association between childhood maltreatment and empathy, with the mean age predicting a reduced negative association between childhood maltreatment and empathy (β = 0.004). However, the percentage of females did not moderate the observed association, which may indicate consistency in the association between childhood maltreatment and empathy across genders.

The results supported the attachment theory, the facial feedback hypothesis, the "like-me" hypothesis, the perception-action model, and the Russian doll model, suggesting that the association between childhood maltreatment and empathy was complicated. These findings not only deepen our understanding of the association between childhood maltreatment and empathy but also produce meaningful practical implications for future research and intervention program design regarding how to promote the development of empathy.

Key words: childhood maltreatment, empathy, perspective taking, personal distress, meta-analysis