Child and adult homicide in Sweden : epidemiological and forensic features

Abstract

Homicide occurs worldwide and is often regarded a major public-health concern. The studies comprising the present thesis were conducted with a view to advancing knowledge of both child and adult homicide from the viewpoints of different branches of forensic science.

In study I, forensic-toxicological data for the years 2007–2009 were retrieved for 120 perpetrators and 265 victims of homicide. Ethanol was the most commonly detected substance, followed by benzodiazepines, which were significantly more common in offenders than victims. Also, perpetrators of homicide–suicide significantly more often displayed negative toxicology than non-suicide homicide offenders. Moreover, victims in unsolved cases were significantly more often positive for narcotics than victims in solved cases. In studies II and III, 200 adult homicide victims and 105 adult homicide offenders were compared by means of logistic regression with 1629 (study II) and 1643 (study III) controls who had lost their lives in vehicle accidents. We found (study II) that ethanol conferred 2-fold and 4-fold risks of homicide victimization and offending, respectively; after stratification by sex, risk estimates were about 3-fold greater in females than males for both homicide offending and victimization. In study III, by comparing prescription data with toxicology data, we found that discontinued treatment with antidepressants conferred a significant 6-fold risk of homicide offending, but no risk increase with regard to homicide victimization; with regard to discontinued treatment with antipsychotics and mood stabilizers, a 7-fold risk was found for homicide offending. For GABA-ergic hypnotics, 2–5-fold risks were found for homicide offending and victimization irrespective of whether treatment had been prescribed or not. In study IV, a range of characteristics from intra- and extra-familial child homicides during 1992 through 2012 are reported. Using Poisson regression, an average yearly decline of 4% in child homicides was uncovered. Substance misuse was found to be uncommon (8%) among child homicide offenders. Prior violent offenses were more common among perpetrators of filicide than filicide–suicide (18% versus 7%); and about 20% of offenders in each group had previously received psychiatric inpatient care. Surprisingly, a third of the extra-familial offenders fulfilled diagnostic criteria for an autism-spectrum disorder. In study V, using a matched case-control design, we found that prior inpatient care conferred a significantly elevated 5-fold risk of filicide victimization. The result was driven by a particularly high risk in females, as no significant risk increase was found in males after stratification by sex.

In conclusion, toxicological findings indicate that ethanol and benzodiazepines are common in homicide offenders and victims; increased risks of homicidal outcome conferred by alcohol were uncovered, as well as increased risks of homicide offending conferred by the use of GABA-ergic hypnotics and the discontinued use of other psychotropic medications. Finally, rates and characteristics of child-homicide offenders and victims were reported, along with risks of filicide victimization conferred by prior healthcare use. Ideally, some of our results might be implemented in everyday healthcare practice and also be useful for homicide investigators and offender profilers.