A cross-sectional study examining relationships between perceived family Expressed Emotion and shame, emotional involvement, depression, anxiety, stress and non-suicidal self-injury, in 264 community and online adults (21.6% male). We compared self-injurers with non-self-injurers, and current with past self-injurers. Self-injurers experienced more family Expressed Emotion (EE) than non-injurers (t(254) = −3.24, p = 0.001), linear contrasts explaining 6% of between-groups variability (F(2, 254) = 7.36, p = 0.001, η² = 0.06). Differences in EE between current and past self-injurers were not significant. Overall shame accounted for 33% of between-groups variance (F(2, 252) = 61.99, p < 0.001, η² = 0.33), with linear contrasts indicating self-injurers experienced higher levels compared to non-injurers (t(252) = −8.23, p < 0.001). Current self-injurers reported higher overall shame than past self-injurers (t(252) = 6.78, p < 0.001). In further logistic regression, emotional involvement and overall shame were the only significant predictors of self-injury status. With every one-unit increase in emotional involvement, odds of currently engaging in self-injury decreased by a factor of 0.860. Conversely, a one-unit increase in overall shame was associated with an increase in the odds of being a current self-injurer by a factor of 1.05. The findings have important treatment implications for engaging key family members in intervention and prevention efforts.
Keywords: Shame; adults; anxiety; depression; emotion dysregulation; expressed emotion; non-suicidal self-injury; stress.