Identifying Cognitive-Affective Risk Subgroups for Explosive Anger in a General Population Sample: A Cluster Analysis of Aggression and Violence-Related Outcomes
Year: 2026
Type: article
Study design: Observational Study
Abstract: Individuals with recurrent impulsive aggression vary considerably in their underlying risk mechanisms. Current approaches to assessment and intervention rarely account for this heterogeneity. Identifying distinct mechanistic profiles within the broader landscape of explosive anger could enable more precise risk assessment and better matching of individuals to targeted treatments. Using a general population sample, this study aimed to identify subgroups based on cognitive, emotional, and impulsivity-related risk factors, and to examine whether these subgroups differ in aggression-related and violence-related outcomes. A UK-based adult general population sample (N = 773; mean age 45.6 years) completed validated measures of cognitive distortions, emotion dysregulation, and trait impulsivity. Cluster analysis (k-means) was used to identify subgroups with distinct mechanistic profiles. The stability and replicability of the clustering solution were evaluated using multiple validation procedures. Subgroups were then compared on aggression history, trait aggression, antisocial behaviour and violent behavioural tendencies. Three distinct risk profiles emerged with differing implications for aggression-related risk. The Multi-Domain High Risk group showed pronounced elevations across all three mechanisms and reported the highest levels of past aggressive behaviour, aggression, and violence. The Elevated-Impulsivity group was characterised primarily by heightened impulsivity with moderate cognitive and emotional dysfunction. The Low-Risk group showed minimal elevations across all mechanisms. Compared to the Low-Risk group, individuals in the Multi-Domain High Risk group were over twice as likely to report a history of violence and three times more likely to endorse violent responses in hypothetical scenarios. Effect sizes for group differences were large (η 2 = .24–.42). Mechanism-based clustering identified interpretable risk profiles that differed substantially in aggression- and violence-related outcomes. These findings provide preliminary support for profiling cognitive, emotional, and impulsivity-related vulnerabilities to characterise heterogeneity in explosive anger within community samples. Longitudinal and clinically diagnosed samples are needed before these profiles can be used to guide clinical assessment or intervention selection.
Source: Psychological Reports
Authors: Fangqing Liu, Charlotte Lennox, Sarah Leonard
Institution: University of Manchester
Language: English
Subfield: Clinical Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: hybrid
Estimated APC paid: $3,650
OpenAlex ID: W7170068134