Edinburgh Research Explorer

The relationship between insight and violence in psychosis: A systematic literature review

Research output: Contribution to journalArticlepeer-review

Related Edinburgh Organisations

Open Access permissions



  • Download as Adobe PDF

    Rights statement: This is an Accepted Manuscript of an article published by Taylor & Francis in The Journal of Forensic Psychiatry & Psychology on 27 December 2019, available online: https://www.tandfonline.com/doi/full/10.1080/14789949.2019.1706760

    Accepted author manuscript, 774 KB, PDF document

Original languageEnglish
Number of pages39
JournalJournal of Forensic Psychiatry & Psychology
Early online date27 Dec 2019
Publication statusE-pub ahead of print - 27 Dec 2019


Poor insight is a risk factor for violence within established risk-assessment tools, yet its relationship to violence in people experiencing psychosis is unclear. To clarify this issue, we sought to systematically review studies investigating the relationship between poor insight and violence in psychosis. A systematic search of studies published between 1980 and 2019 was carried out using Pubmed, Embase, Medline, PsychInfo and CINAHL databases. From combined search results of 5701 articles, 18 observational studies met the inclusion criteria and were selected for full-text review and quality grading. Eight demonstrated a positive relationship between poor insight and violence whilst 10 failed to find this relationship. Significant methodological limitations were found across studies. Those measuring the clinical insight dimension specifically and reliably were most able to demonstrate a positive relationship between poor insight and violence. Choice of the measurement tool and co-variates such as psychopathy were found to influence this relationship. We, therefore, found partial evidence in support of a relationship between poor insight and violence in psychosis. In order to gain an enhanced understanding of this relationship, better quality research accounting for relevant co-variates and using appropriate measurement tools which target the ‘clinical’ insight dimension is required.

    Research areas

  • psychosis, insight, violence, schizophrenia, aggression

Download statistics

No data available

ID: 130132728