S Strand1, T E McEwan. 1. Mid-Sweden University and Sundsvall Forensic Psychiatric Hospital, Sundsvall, Sweden.
Abstract
BACKGROUND: Female stalkers account for 10-25% of all stalking cases, yet little is known about risk factors for female stalking violence. This study identifies risk factors for female stalking violence and contrasts these with risk factors for male stalking violence. METHOD: Seventy-one female and 479 male stalkers presenting to police in Sweden and a specialist stalking clinic in Australia were investigated. Univariate comparisons of behaviour by gender, and comparisons between violent and non-violent female stalkers, were undertaken. Logistic regression was then used to develop a predictive model for stalking violence based on demographic, offence and clinical characteristics. RESULTS: Rates of violence were not significantly different between genders (31% of males and 23% of females). For both men and women, violence was associated with a combination of a prior intimate relationship with the victim, threats and approach behaviour. This model produced receiver operating characteristic (ROC) curves with area under the curve (AUC)=0.80 for female stalkers and AUC=0.78 for male stalkers. The most notable gender difference was significantly higher rates of personality disorder among women. High rates of psychotic disorder were found in both genders. Stalking violence was directly related to psychotic symptoms for a small number of women. CONCLUSIONS: Similar risk factors generally predict stalking violence between genders, providing initial support for a similar approach to risk assessment for all stalkers. The most notable gender difference was the prevalence of personality and psychotic disorders among female stalkers, supporting an argument for routine psychiatric assessment of women charged with stalking.
BACKGROUND: Female stalkers account for 10-25% of all stalking cases, yet little is known about risk factors for female stalking violence. This study identifies risk factors for female stalking violence and contrasts these with risk factors for male stalking violence. METHOD: Seventy-one female and 479 male stalkers presenting to police in Sweden and a specialist stalking clinic in Australia were investigated. Univariate comparisons of behaviour by gender, and comparisons between violent and non-violent female stalkers, were undertaken. Logistic regression was then used to develop a predictive model for stalking violence based on demographic, offence and clinical characteristics. RESULTS: Rates of violence were not significantly different between genders (31% of males and 23% of females). For both men and women, violence was associated with a combination of a prior intimate relationship with the victim, threats and approach behaviour. This model produced receiver operating characteristic (ROC) curves with area under the curve (AUC)=0.80 for female stalkers and AUC=0.78 for male stalkers. The most notable gender difference was significantly higher rates of personality disorder among women. High rates of psychotic disorder were found in both genders. Stalking violence was directly related to psychotic symptoms for a small number of women. CONCLUSIONS: Similar risk factors generally predict stalking violence between genders, providing initial support for a similar approach to risk assessment for all stalkers. The most notable gender difference was the prevalence of personality and psychotic disorders among female stalkers, supporting an argument for routine psychiatric assessment of women charged with stalking.