Abstract
Research on emotion regulation (ER) and aggression has increased rapidly in the past years, but heterogeneity across studies make integration of findings challenging. To estimate the size and consistency of the relationship between ER and aggression we conducted a systematic review and meta-analysis. In total, 16,369 articles were retrieved from Scopus, PubMed, PsycINFO, and CINAHL, and 137 articles (171 studies) provided 918 effect sizes from a total sample of N = 252,605. Multilevel models showed significant pooled relationships, with small-to-moderate effects. Use of adaptive strategies was associated with lower aggression (r = -0.090, I(2) = 78%), but the precision estimate test (PET) bias corrected effect was non-significant. Use of maladaptive strategies (r = 0.329, I(2) = 93%), and difficulties regulating emotions (r = 0.248, I(2) = 89%), were associated with higher aggression. P-curves suggested little evidence of selective reporting while cumulative meta-analysis showed relatively consistent effects over time. Moderator analyses showed that effects were generally stronger for physical and other forms of aggression than sexual aggression, and weaker for intimate partner targets than other targets. Difficulties in ER were most strongly related to reactive compared with proactive aggression, but the use of ER strategies did not differ between motivations. Our findings suggest that interventions targeting ER, including those that explore and challenge angry predispositions and provide skills to manage impulse control, may help to prevent aggressive behaviour.