When Empathy Helps and Hurts

Non-Linear Associations Between Early Childhood Maltreatment History, Empathy and Self-Harm Thoughts

Author

Andy Grogan-Kaylor

Published

February 9, 2026

⬇️

Objective

Child maltreatment is a well-established risk factor for self-harm, yet the affective processes underlying this association remain unclear.

This study (Hong et al., 2026) examines how empathy relates to the association between early childhood maltreatment and lifetime self-harm thoughts.

Method

Data from 303 adults were stratified based on the presence or absence of maltreatment history in early childhood. Self-reports of lifetime self-harm thoughts in adulthood and empathy during adolescence were examined in both groups using generalized additive models (Wood, 2017, 2025) to assess non-linear relationships. Models also controlled for respondents’ age, gender, and socio-economic status.

Results

There was an interesting and complex pattern of results.

Adults with early childhood maltreatment history were more likely to report self-harm thoughts than those without such a history.

A significant nonlinear association was observed between empathy and self-harm thoughts.

Predicted probabilities of self-harm thoughts were lowest at moderate levels of empathy for both groups.

Risk for self-harm thoughts was highest among individuals with high empathy.

Risk for self-harm thoughts was also somewhat high for individuals with low empathy.

Conclusion

While a moderate level of empathy can be protective from early childhood maltreatment, both low and high levels of empathy were associated with a greater risk for self-harm thoughts.

Findings suggest the need to further investigate the role of empathy as a “risky strength” among adults.

Future Research

While the temporal ordering of the variables in this study supports the idea of empathy as a “risky strength”, reverse causality is also a possibility. It may be that different levels of empathy are an outcome of different levels of self-harm thoughts. Future research would do well to explore these questions with larger samples, especially those that have repeated measures of the constructs of interest, and advanced quantitative methods capable of offering more causally robust conclusions.

flowchart LR
  
  subgraph childhood
  
  m("maltreatment")
  
  end
  
  subgraph adolescence
  
  e("empathy")
  
  end
  
  subgraph adulthood
  
  sh("self harm thoughts")
  
  end

  m --> sh

  e --> sh
  
  childhood ~~~ adolescence
  
  adolescence ~~~ adulthood

  style m fill:#9A3324,stroke:#000000,stroke-width:2px,color:#FFFFFF;

  style e fill:#374EA2,stroke:#000000,stroke-width:2px,color:#FFFFFF;

  style sh fill:#FFC20E,stroke:#000000,stroke-width:2px,color:#000000;

  style childhood fill:#D9D9D9,stroke:#000000,stroke-width:2px,color:#000000;

  style adolescence fill:#D9D9D9,stroke:#000000,stroke-width:2px,color:#000000;

  style adulthood fill:#D9D9D9,stroke:#000000,stroke-width:2px,color:#000000;
  
  linkStyle 0,1 stroke:#000000,stroke-width:3px,font-size:36px,color:black;



flowchart TB
  
  subgraph maltreatment
  
  direction LR
  
  e1("empathy") --> sh1("self harm thoughts")
  
  end
  
  subgraph nomaltreatment ["no maltreatment"]

  direction LR
  
  e0("empathy") --> sh0("self harm thoughts") 
  
  end
  
  style e0 fill:#374EA2,stroke:#000000,stroke-width:2px,color:#FFFFFF;

  style e1 fill:#374EA2,stroke:#000000,stroke-width:2px,color:#FFFFFF;

  style sh0 fill:#FFC20E,stroke:#000000,stroke-width:2px,color:#000000;
  
  style sh1 fill:#FFC20E,stroke:#000000,stroke-width:2px,color:#000000;

  style maltreatment fill:#9A3324,stroke:#000000,stroke-width:2px,color:#FFFFFF;
  
  style nomaltreatment fill:#D9D9D9,stroke:#000000,stroke-width:2px,color:#000000;

  linkStyle 0,1 stroke:#000000,stroke-width:3px,font-size:36px,color:black;

overall results

maltreatment

empathy

References

Hong, S., Grogan-Kaylor, A., Kim, M., Bender, A. E., Saba, S., Yu, C.-L., Fedina, L., & Herrenkohl, T. I. (2026). When empathy helps and hurts: Non-linear associations between early childhood maltreatment history, empathy and self-harm thoughts. Child Maltreatment. https://doi.org/10.1177/10775595261422448
Wood, S. N. (2017). Generalized additive models: An introduction with R (Second edition.). CRC Press, Taylor & Francis Group.
Wood, S. N. (2025). Generalized additive models. Annual Review of Statistics and Its Application, 12, 497–526. https://doi.org/10.1146/annurev-statistics-112723-034249