We have so much in common: Does shared variance between emotion-related constructs account for relationships with self-injury?

Abstract

Background: Emotion regulation, distress tolerance, experiential avoidance, and both positive and negative affect have all been linked to NSSI. These constructs are proposed to be distinct; however, they share conceptual similarities. For example, some people may regulate emotions by avoiding stressful situations, conflating emotion regulation and avoidance. We tested if constructs linked with NSSI (when studied in isolation), remain significant correlates of NSSI when considered alongside related constructs (with which they may share variance). Method: University students (n = 487, M = 21.36, SD = 2.48, 74% female, 40% with lived experience of self-injury) completed well-validated self-report measures of NSSI, difficulties with emotion regulation, distress tolerance, experiential avoidance, emotional reactivity, positive and negative affect, and alexithymia. Results: As predicted, emotion-related constructs were generally highly correlated. Additionally, with the exception of lack of emotional awareness, all constructs were significantly associated with NSSI in bivariate analyses. In multivariate analyses, associations were substantially attenuated. Positive affect, distress tolerance, and experiential avoidance were negatively associated with NSSI, and limited emotion regulation strategies was positively associated with NSSI. No other constructs were uniquely associated with NSSI and exploratory factor analyses indicated that all constructs loaded onto a single factor Limitations: Cross-sectional design rules out temporal sequencing. Conclusion: Findings raise the possibility that associations between some emotion-related constructs (e.g., alexithymia) and NSSI may reflect variance shared with other emotion-related constructs. If true, this will have important theoretical, clinical, and measurement implications for NSSI research.