ORIGINAL PAPER Min Yang Jeremy Coid Gender differences in psychiatric morbidity and violent behaviour among a household population in Great Britain Received: 2 October 2006 / Accepted: 21 May 2007 / Published online: 27 June 2007 j Abstract Background Men are more violent than women. It is unclear whether psychiatric morbidity contributes to this gender difference in the general population. This study examined gender differences in psychiatric diagnosis and violent behaviour to test whether risk from psychiatric morbidity accounted for gender differences in violent behaviour; whether violent males were more vulnerable to risk from psychiatric morbidity; and, whether violent women surmounted a higher threshold of risk from psychi- atric morbidity. Methods Cross-sectional random sample of 8,397 household residents in Great Britain was assessed using structured questionnaires to measure violent behaviour and psychiatric morbidity. Weighted multilevel regression models were used to test hypotheses, adjusting for demographic factors, psychiatric comorbidity and area effects of violence. Results Men were more likely to report violence than women and were exposed to greater risks from sub- stance dependence, Antisocial Personality Disorder (ASPD) and hazardous drinking. Women were ex- posed to greater risks from affective/anxiety disor- ders. Psychiatric morbidity explained 22% of the gender difference in violence. Violent men were less vulnerable to risk from any Personality Disorder (PD), in particular ASPD, than violent women. Vio- lent women reported a higher etiological threshold for affective/anxiety disorders and any PD; violent men higher threshold for alcohol dependence and haz- ardous drinking. Conclusions Psychiatric morbidity has moderate impact on higher levels of violence among men. Antisocial Personality Disorder poses a greater risk for violence among women than men. Affective/anxiety disorders and any PD are more severe conditions among violent women; alcohol dependence and hazardous drinking are more severe among violent men, confirming the ‘‘Threshold of Risk’’ hypothesis. j Key words national survey – violent behaviour – psychiatric morbidity – gender differences Introduction Epidemiological surveys have demonstrated strong associations between psychiatric morbidity and reported violent behaviour [15, 27, 28] and that pre- valences of psychiatric disorders differ between men and women [12, 17]. Gender differences between different forms of violence are reported in samples of young couples [16], psychiatric inpatients [7, 13] and mentally ill offenders [30]. It is possible that differ- ential levels of risk associated with specific diagnostic categories could explain why males are generally more antisocial than females [10, 21]. Research into gender-differentiated effects of risk factors has been stimulated by the threshold of risk (gender paradox, or group resistance) hypothesis, developed and ap- plied to study the etiology of antisocial disorder [5, 23]. This hypothesis argues that females who develop antisocial behaviour have a higher threshold of risk than males and are therefore more severely afflicted. Research supporting this hypothesis has demon- strated that anxiety disorder increases the risk of victimization to a greater extent among young women than among young men [16], whilst substance abuse increases the risk of violence to a greater extent among males [13, 16]. More recently, a longitudinal study of young per- sons in Dunedin, New Zealand, has attempted to ex- plain gender differences in antisocial behaviour by Dr. M. Yang (&) J. Coid Forensic Psychiatry Research Unit St. Bartholomew’s Hospital, William Harvey House 61 Bartholomew Close London EC1A 7BE, UK Tel.: +44-20/7601-7511 ext - 2171 Fax: +44-20/7601-7969 E-Mail: m.yang@qmul.ac.uk Soc Psychiatry Psychiatr Epidemiol (2007) 42:599–605 DOI 10.1007/s00127-007-0226-8 SPPE 226