Violent crime is one of the leading causes of reduced health, disability, and death worldwide, not least among young people. Forensic psychiatry is striving to identify mental disorders as contributing causes behind violent crimes. In high-quality epidemiological studies, the major mental disorders, such as schizophrenia and mood disorders, have been found to carry only a very modest risk-increase for violent offending, while the vast majority of individuals with these disorders are as law-abiding as everyone else. Substance-related disorders are overrepresented among violent offenders, but the main culprit among mental disorders for aggressive antisociality may be the childhood-onset disruptive behaviour disorders. The scientific study of these disorders has been hampered by convoluted and overlapping diagnostic definitions. Conduct disorder (CD) is comparatively easy to grasp, as it directly reflects criminal and violent behaviours in childhood and adolescence. Antisocial personality disorder (ASPD) is CD continued into adulthood with the addition of symptoms reflecting emotional and interpersonal dysfunctions (DSM-IV, APA 1994). Finally, psychopathy has been proposed to constitute a personality disorder independent of violent and criminal behaviour, albeit a strong risk factor for it. Childhood-onset hyperactivity and impulsivity have been discerned in the background to all these conditions, while autistic-like restrictions in social cognition and emotional processing may be an important co-factor. In the present project, persistent violent behaviours will be studied as dependent variable per se (in line with the CD definition in the DSM-IV, conferring considerable advantage in clarity over mixed definitions such as “psychopathy”) to identify and quantify psychiatric, genetic, and environmental susceptibility factors.
The specific aims of the project are numbered by the three specific sub-studies presented below:
1. to identify epidemiological risk factors for persistent violent crimes in the general population,
2. a) to quantify the roles of childhood neuropsychiatric problems, including attention-deficit/hyperactivity disorder (AD/HD) and autism spectrum disorders (ASDs), in the development of CD, ASPD, and persistent violent crimes,
b) to quantify the influence of common and specific genetic and environmental susceptibility factors for this development,
3. to identify psychiatric disorders and personality traits that contribute to persistent violent crimes among offender groups.
Since its early days, psychiatry has been applied to societal concerns, especially crime and punishment, which eventually resulted in the establishment of forensic psychiatry as a specific specialty. Though sometimes the target for vehement criticism, forensic psychiatry has gained considerable influence on, for example, the current Swedish penal law (“Brottsbalken”) and court proceedings related to violent crimes. Sweden has been one of the leading countries in forensic psychiatric research, especially by applying neurobiological (e.g. Lidberg et al., 1984) and epidemiological (e.g. Grann & Fazel, 2004) methods. The official systems for expert opinion and treatment are well developed. There is, however, a vast discrepancy between the demands on forensic psychiatry to identify psychiatric susceptibility for crimes, predict the likelihood of their recurrence in individuals, and treat patients in order to prevent future crimes on the one hand, and, on the other, the increasing demands for a reliable evidence basis in medical practice. The main diagnostic schemes applied in forensic psychiatry have neither been specifically validated for the purposes they are used for, nor tested among the relevant patient categories. Research has been impeded by a lack of clarity in goals and dependent variables: is it mental disorders or the propensity to commit crimes that is being studied and treated? Even if patient groups differ in core aspects, evidence for the efficacy of treatment methods is most often extrapolated from general psychiatry. In addition to these problems, forensic psychiatry, unlike most medical specialties, has to adopt a national perspective, as its practice and patients are determined by the current legislation and current sociodemographic, economic, and cultural contextual factors, while international research has to be “translated” and adapted to the Swedish situation.
This relative lack of evidence-based medical methods for problem detection and prevention stands in sharp contrast to the health impact of violence. Violence accounts for about 14 % of deaths among males aged 15-44 years and for 7 % of deaths among females in the same age group. A murder is likely to severely disrupt the lives of numerous individuals besides the victim and the perpetrator, e.g. families, friends, acquaintances, witnesses, suspects, and others. Also violent crimes with less devastating consequences, such as assaults, have the potential to inflict widespread damage. Further, fear of violence and the security measures necessary to relieve public anxiety constitute huge psychological and economic costs to society. Preliminary studies for the present project show that violent crimes are perpetrated by a small group of humans in society (4.5 % of adult Swedes are at some time sentenced for a violent crime), and that an even smaller group is responsible for a disproportionate number of crimes (>60 % of all violent crimes are committed by less than 1 % of the population).
The most promising psychiatric candidates to explain both the propensity to violent behaviours and the type of substance abuse that is linked to violence are to be sought among definitions of childhood behavioural problems (Robins 1966, Cloninger et al., 1981). In the DSM-IV system, childhood disruptive behavioural disorders include CD, oppositional defiant disorder (ODD), and AD/HD. The nosological status of ODD remains disputed as it is found in the vast majority of children with AD/HD (Biederman et al., 1996). AD/HD or other neurodevelopmental problems are virtually a prerequisite for pre-pubertal CD, while later-onset CD may develop from various psychological or social problem constellations (Lahey & Loeber, 1997). About a third of children with AD/HD go on to have CD, a combination that carries a discouraging prognosis with high risks for substance abuse, criminality, mental disorders, and early death. Children with AD/HD who do not develop childhood CD face a much better future, almost on par with the children who do not have a mental disorder (Hofvander et al., 2009).
Autism spectrum disorders (ASDs, i.e. autism, Asperger syndrome, atypical autism, and autistic traits, referred to as ‘pervasive developmental disorders’ in the DSM system) form a cluster of social interaction behavioural disorders. The majority of all children with AD/HD also display several symptoms of DSM-IV autistic disorder, such as social interaction or communication deficits, and an important minority meet full diagnostic criteria for an ASD (Kadesjo & Gillberg, 2001; Sturm et al., 2004; Gadow et al., 2006). It is still unclear how this overlap should best be accounted for in diagnostics and whether social interaction and communication problems, such as those defined on the autism spectrum, are associated with adult ASPD and persistent violent behaviours. Indications pointing in this direction have been presented, however, both in studies of adolescents and retrospective studies in adults focusing on empathy problems or limited capacity for emotional processing (Siponmaa et al., 2001; Gilmour et al., 2004; Anckarsäter et al., 2007).
Childhood behavioural disorders are also important background factors to other major adult psychiatric disorders, such as schizophrenia (Vourdas et al., 2003) and bipolar disorder (Egeland et al., 2003). Semantic-pragmatic language distortions that are well described in childhood disorders (Noens & van Berckelaer-Onnes, 2005) are not included in adult personality disorder definitions despite the “semantic blindness” once proposed to constitute the core nucleus of psychopathy (Cleckley, 1941), and the deviant emotional processing, with abnormalities in both appraisal and use of emotional information, reported from various clinical populations (Parrott et al., 2005). In ASD and schizophrenia alike, specific aberrations in the processing of abstract content and metacognition have been proposed as core disordered processes (Koren et al., 2004).
Heredity factors account for the lion’s share of the variation in ASD or severe AD/HD, and are more important for childhood-onset, life-course-persistent antisocial behaviour than for adolescent-limited criminality (Eley et al., 2003). Genetic factors, however, may have a lower explanatory value for problems in severely under-privileged social settings (Tuvblad et al., 2006). Genetic factors have also been reported to contribute both to stability and change in AD/HD symptoms (Larsson et al., 2004). By using the childhood-onset phenotypes to tap into neurocognitive functioning, such as AD/HD and executive functions or ASD and empathy, the understanding of genetic mechanisms involved in adult personality may be advanced (e.g. Anckarsäter et al., 2006), and these early manifestations of deviant neurodevelopment and cognition hold promise as endophenotypes in psychiatric molecular genetic research.
In this project, psychiatric background factors to persistent violent behaviours will be studied in an epidemiological database of the general Swedish population (sub-project 1), a nation-wide longitudinal twin study of adolescent development (sub-project 2), and several longitudinal clinical studies (sub-project 3). A reasonably small number of phenotype definitions (persistent violent offending, CD, and ASPD) will be assessed as dependent variables and the degree to which they may be explained by psychiatric factors (including the childhood-onset behavioural disorders) quantified in multivariate models, also accounting for various psychosocial and economic co-variates. The three sub-projects follow the specified Aims of the project.
1. Epidemiological characteristics of individuals responsible for persistent violent crimes
Research questions: 1. Which are the basic characteristics (gender, age, age at onset, schooling, prevalence of mental disorders and hereditary and birth negative factors) of the population group responsible for the majority of violent crimes?
2. Which psychiatric factors predict persistent violent offending when socioeconomic factors, early onset of criminality and school failure are controlled for?
Description: According to official Swedish statistics and preliminary analyses of our data, the majority of violent crimes are perpetrated by persons who have earlier convictions for such crimes. While it will probably always remain difficult to predict first offences, it will be more feasible to characterize the group of persistent offenders in order to assess the predictability of recidivistics crimes. A de-identified data-base has been created by a linkage between the Multi-Generation Register and the National Crime Register, the Client Register of the National Board of Institutional Care, the Register of Education, the Population and Housing Census, the School Register and the Migration records from Statistics Sweden, the Cause of Death register, the Birth Register, the Hospital discharge registry, the Military Archives of Sweden and the Swedish Twin Registry.
Subjects and methods: Among persons born between 1958 and 1980 and living in Sweden on their 15th birthday (n=2 401 424), four groups are identified. a) the smallest subgroup responsible for the majority of violent crimes, b) the group responsible for all violent crimes, c) the group responsible for all other types of crimes, and d) non-criminal population controls, including sibling-controls to match for possible confounding factors. These groups will be compared for the following independent variables: age at onset, sex, socio-economic stratum (of parents and proband), family history (of criminality, mental disorders, substance abuse, suicide attempts and early death), birth complications, school grades, conscription results, early death, neurological diagnoses and all types of mental disorders. Multivariate logistic regression models will be used to identify predictors of persistent violent crimes (as compared to single-event violent offending, general criminality, or no criminality). The “goodness of fit” of the final predictive model will be assessed and explored to quantify the degree to which the predictor variables explain the defined groups. The study cohort is the largest possible and has statistical power for the planned analyses.
External and internal collaboration: This study is a collaboration with Professors Lichtenstein and Långström at Karolinska Institutet, Department of Medical Epidemiology and Biostatistics, who have created the database and will co-supervise the analytical work which will be performed by Örjan Falk, BA, medical student, with Nóra Kerekes, PhD, as tutor.
2. The Child and Adolescent Twin Study in Sweden (CATSS): the Conduct Disorder study
Research questions: 1. Which childhood mental health problems may be linked to the development of persistent violent criminality and how large is their contribution in interaction with environmental susceptibilities?
2. Are there common (in addition to specific) genetic and environmental susceptibilities behind persistent violent criminality and mental health problems?
Description: The Child and Adolescent Twin Study in Sweden (CATSS) is a longitudinal study including all Swedish birth cohorts of twins born since 1992, the first of whom thus turning 18 in 2010. Mental and somatic health, together with a range of psychosocial and economic data, has previously been assessed at 9, 12, and 15 years of age. Using the longitudinal, multi-source measure of violent DSM-IV CD at 18 years of age as dependent variable, background factors will be sought among the assessments of mental health problems at 9 or 12 years of age.
Subjects and methods: Since 2004, parents of all 9- and 12-year-old twins in Sweden are interviewed over the phone (CATSS-9/12, current n >18.000, >80 % participation rate), followed by new waves of questionnaires to the parents and the twins at 15 (CATSS-15) and 18 (CATSS-18) years of age. The base-line telephone interviews at age 9 or 12 include the Autism, Tics, AD/HD and other Comorbidities Inventory (A-TAC), covering child neurodevelopmental and other psychiatric problems, validated in several phases (Hansson et al., 2005, Larson et al., 2010, Halleröd et al., 2010). For the first six year cohorts, a parent inventory of psychopathic personality traits is available. Socio-economic and family data are collected at base-line. In a substudy of the twins who are screen-positive for ASDs, AD/HD, tics, learning disorders, coordination disorder, and/or CD, questionnaires measuring life-events, parent-child relationships, child monitoring, school environment, and dyadic adjustments have been collected. DNA is currently collected from all twins for molecular genetic analyses.
At 15 years of age (CATSS 15), both the parents and the twins are given questionnaires measuring, among other variables, aspects of friendship, personality, criminality, alcohol and drug use/abuse, bullying, physical activity, family environment, societal measures (e.g. foster care, institution placement) asthma, allergy, and puberty. In addition, questionnaires tapping into personality development and prodromal phases of schizophrenia and bipolar disorder are used. An in-depth clinical study is performed for birth cohorts from the 1993, started in 2008 (this study is detailed in the application by Dr Gumpert to the Research Council this year). At 18 years of age, the CATSS-18, which will start this summer, includes measures of self reported criminality, alcohol and drug use/abuse, personality (including ASPD), neurodevelopmental problems, school performances, and life-events. The 18-year follow-up data will be used to measure the dependent variables (self-reported violent crimes and CD) for the study presented here. The present application refers to the collection of the CD phenotypes at age 18 during the upcoming three years and predictive analyses using CD at 18 as dependent variable and A-TAC and psychopathic personality scores together with psychosocial and economic data from the previous assessments as independent predictors in linear and logistic regression models with generalized estimation equations to account for the dependency within twin-pairs. In analyses where both the exposure and the outcome are continuous, mixed-effect models will be fitted. As the study builds on nation-wide cohorts, statistical power cannot guide the data collection but will be ascertained pre-hoc for the individual analyses.
Preliminary results: Neurodevelopmental problems are dimensionally distributed in the population, strongly inter-correlated, and form factors reminiscent of adult personality structure (Anckarsäter et al., 2008). Papers on the development and validation of the A-TAC instrument have been published, and ten papers on cross-sectional data from the CATSS 9/12 are currently in press, submitted, or in final preparation (e.g. Lundström et al., 2010). The A-TAC has cross-validity with other measures, such as the junior Temperament and Character Inventory (J-TCI, Kerekes et al., submitted) and the Child Behaviour Check-List (CBCL, Halleröd et al., 2010). Twin analyses are underway in several national and international collaborative networks (Ronald et al., submitted, Losch et al., submitted).
External and internal collaboration: The CATSS is a collaborative project with Professor Lichtenstein at Karolinska Institutet. Other projects outgoing from the CATSS are headed by us in collaboration with Professors Råstam and Bölte and Dr Borg (autism), Dr Gumpert (psychosocial marginalization and mental health problems in adolescence), Professor Långström (development at 15 years of age) and MSc Lundström (development at 18 years of age). PLEASE NOTE THE DELINEATION BETWEEN THE PRESENT APPLICATION (CD PHENOTYPE, longitudinal CATSS) FROM THE APPLICATIONS SUBMITTED BY RÅSTAM (AUTISM/ADHD, clinical follow-up at 9 years), BÖLTE (AUTISM, clinical follow-up at 18 years), BORG (AUTISM, PET methods) AND GUMPERT (ADOLESCENT PSYCHOSOCIAL ADAPTATION, clinical follow-up at 15 years). THESE FOUR APPLICATIONS ALL BUILD ON THE CATSS COHORTS BUT EACH ENTAILS NEW DATA COLLECTIONS USING SPECIFIC METHODS IN DIAGNOSTIC SUBGROUPS AND THUS DO NOT OVERLAP WITH THE PRESENT APPLICATION IN AIMS, QUESTIONS, OR DATA COLLECTED. Sebastian Lundström, PhD candidate, works exclusively with the project and will defend his PhD thesis next academic year, while Tomas Larson and Sara Lina Hansson Halleröd have worked with the validation of the A-TAC and will write their theses on diagnostics. Robert Sandin is a medical student working on the overlap between disruptive behaviour disorders and asthma/allergies. Sven Brändström is a postdoctoral fellow who works exclusively with the longitudinal development of personality according to the TCI model in relation to neurodevelopmental disorders. Nóra Kerekes is a postdoctoral fellow who will be responsible for the data analyses presented in the present application.
3. Upprätt (Uppföljningsstudier i rättspsykiatri/Forensic psychiatric follow-up studies)
1. Are there adult psychiatric symptom clusters/disorders and personality traits that predict persistent violent crime in the presence of CD, substance abuse, and previous criminal records?
2. Are there specific predictors of persistent violent crimes among patients treated in forensic psychiatry as compared to subjects serving time in the general criminal justice system?
Description: Several study groups and cohorts (most of which are previously reported in cross-sectional papers) are prospectively followed to study the relation between mental disorders and persistent violent crimes. The goal is to define the adult phenotype most strongly related to the majority of violent crimes in society.
Subjects and Methods: All study groups and cohorts (listed below) were assessed by similar or compatible research protocols at baseline. Psychiatric disorders, including childhood-onset neuropsychiatric disorders, were diagnosed according to the DSM-IV, often by the SCID-I and SCID-II interviews. In some of the study groups, the Autism – Tics, AD/HD and other Comorbidities inventory (A-TAC) was used to further assess childhood psychiatric problems. Personality traits were in most cases assessed by the Karolinska Scales of Personality (KSP) and the Temperament and Character Inventory (TCI), whereas the junior version of the TCI, J-TCI, was used for adolescents. Risk assessments were carried out by means of the Historical, Clinical, and Risk-management 20 (HCR-20) and the Psychopathy Check List-Revised (PCL-R). Neurocognitive abilities were assessed by the Wechsler test battery, WAIS-R for individuals recruited before 2003, and WAIS-III for those recruited after 2003. WISC-III was used for adolescents/individuals under 16 years of age. For most of the study groups, psychosocial data were collected in accordance with a research protocol covering demographic variables, family history, education, and work history as well as circumstances of the index crime and criminal history profiling (age at onset, patterns of criminal behaviours).
Follow-up data is or will be collected from the National Council for Crime Prevention, the National Board of Health and Welfare, and the National Prison and Probation Administration registers for the following outcomes: (i) incidents during treatment (forensic psychiatric/institutional/prison), (ii) discharge, (iii) recidivistic crimes (all types, particularly sexual and violent crimes, according to frequency and severity), (iv) new inpatient psychiatric treatment episodes for substance abuse or major mental disorders, (v) somatic treatment for accidents, suicide attempts, or violence (victimization), (vi) death and causes of death.
The following cohorts/study groups are presently followed as outlined above (publications describing the groups at baseline are referenced or in preparation):
The Göteborg Forensic Neuropsychiatry Project. One hundred perpetrators of severe violent and/or sexual crimes were assessed at the Department of Forensic Psychiatry in Göteborg between 1998 and 2001. DSM Axis I diagnoses of childhood-onset neuropsychiatric disorders, Axis II diagnostics by interviews and self-ratings, psychosocial factors, neurocognitive tests, and risk instruments, such as the PCL-R and the HCR-20, were collected together with neurobiological markers (Soderstrom et al., 2004 and 2005). The first follow-up studies have been published (Gustavson et al., 2010, Wallinius et al., submitted, Nilsson et al., in preparation).
The Göteborg Sexual Child Abuse studies. An epidemiological cohort of all 198 perpetrators sentenced for sexual crimes against minors in the Västra Götaland area between 1993 and 1997 (Carlstedt et al., 2001), completed by a study group of 341 forensic investigees with sexual crimes against minors or adults from the same period (Carlstedt et al., 2005).
The Malmö clinical forensic psychiatric cohort. All subjects sentenced to forensic psychiatric inpatient treatment between 1999 and 2005 and living in the Malmö area (n=135) are included with data on base-line assessments (i.e. the forensic psychiatric reports leading up to the sentences, including a number of structured instruments, such as the SCID I and II, PCL-R and HCR-20), and a detailed time-line for the whole treatment period, including incidents, leaves, permissions, changes of contact persons, medication, psychotherapy, and follow-up data after discharge. The cross-sectional study is in preparation (by Nyman, Andreasson and coworkers).
Clinical, Actuarial, and Treatment-related prognostic factors in SiS institutions (SiS-CAT). A group of 110 consecutive adolescents admitted to one of several institutions within the National Board of Institutional Care were assessed by broad psychiatric, personality, neurocognitive, and social test batteries (including DSM-IV diagnostics, A-TAC, WISC-III or WAIS-III, J-TCI, Beck Youth Inventories, Youth Self-Report (YSR), Youth Psychopathic Inventory (YPI) or the PCL-R) for a prospective clinical study (Ståhlberg et al., preliminarily accepted). Two papers from a pilot clinical evaluation project have been published (Anckarsäter et al., 2007 and 2008).
Two cohorts of all male perpetrators from the Västra Götaland region of Sweden, consecutively admitted for forensic psychiatric investigation between 1988 and 1990 (n=110), and between 1993 and 1995 (n=186), have been used to study the practical implementation of the changes in legislation carried out in 1992 and outcome in different types of sanctions (forensic psychiatric care, prison, other, Lund & Forsman, 2005; Lund, Forsman & Anckarsäter, submitted).
Statistical methods and power: Across these study groups and cohorts, persistent violent offending as dependent variable will be compared to psychiatric and other predictors by non-parametric statistics such as Spearman rho, Mann-Whitney U test, and Kruskal-Wallis H test. The Kaplan-Meier survival method (also known as the product limit estimator) will be used to estimate the survival function from outcome data such as time until criminal reoffending, re-hospitalization, and untimely death. Finally, regression models will be used to identify predictors and to test psychiatric prediction models for relapse into reoffending, substance abuse, re-hospitalization, and untimely death. bivariate models, controlling for previous criminal behaviour, CD, and substance abuse. The amount of variance in outcome measures that is explained by different predictor variables will be estimated. In order to test whether the samples/cohorts are large enough to statistically identify predictors, power analyses have been performed, using the G*Power programme for the multiple regressions. Testing six predictors with an alpha equal to .05, a sample of 100 subjects, and a medium to large effect size will give a power of at least 80 %. The cohorts included in the project at minimum consist of 100 subjects and often exceed this number, giving sufficient power for the planned analyses. The cohorts will be analyzed independently, allowing subsequent analyses of primary hypotheses that will be used when the statistical power in the first cohort does not allow for adjustment of alpha according to the number of initial comparisons.
Significance and Preliminary results: Due to the number and size of the groups compiled, it will be possible to combine conceivable risk factors within these groups of mentally disordered offenders, and by testing risk factors against different outcome measures, true predictors will be separated from mere artifacts without substantial validity. So far, proposed risk factors and instruments have been modestly correlated with violent recidivism on the group level, while they have showed a poor predictive power for the individual. A number of studies assessing predictive models for relapse in criminal behaviours are now headed by Thomas Nilsson, PhD. Three doctoral candidates, Christina Gustavson, Christina Lund, and Märta Wallinius, will defend their theses on this project during the upcoming academic years. The present application specifically refers to the analyses of persisting violent crimes as phenotype across different sanction categories.
Two further aims may be possible to address by this project. First, to identify the empirical phenotypes that best match the genetic susceptibility for AD/HD with early-onset CD and persistent violent criminality (by identifying common genetic effects in twin analyses and subsequently identify and weight the items that are most strongly correlated with this underlying susceptibility by correlations, factor analyses, and measures of internal consistency. Such empirical phenotypes may advance analyses of psychiatric co-variates to findings from molecular genetic research with emerging genome-wide techniques. Second, it will be possible to quantify, at least as a crude approximation, the overall causation of violent crimes (persistent or single-event) by genetic and psychiatric conditions by squared tetrachoric and multivariate correlation coefficients. This analysis will be relevant for the broader issues of risk assessments and attitudes towards offenders.
All studies where so requested have been approved by the relevant ethical review boards, sub-project 1: 2009/939-31/5, for sub-project 2: 03-672 and for the sub-projects in 3: 724-96, S 99-97, L 527-97, L 304-97, S-196-98, L 400-98, Ö 545-01, Ö 034-02, Ö 035-02, Ö 465-02, 193-05, 64/2007. The research group has considerable experience in conducting clinical and epidemiological research on vulnerable groups (Munthe et al, 2010). High security surrounds all work with person data, generally performed on anonymized data files. A strict and well-reflected media policy for the presentation of data and abstaining from comments on individual cases is applied.
Criminal behaviours are predominantly a male phenomenon. Though female offenders differ from males in several respects, women have been included in almost all the studies above, which so far has yielded some specific findings for females in spite of the small study groups (e.g. autism spectrum disorders among incarcerated females). Beyond questions of sex ratios, we have not addressed gender, in the sense of a socially constructed sex identity, but are, of course, interested in possible interpretations in this direction from personality profiles. It is obvious, however, that most medical studies speaking of gender never advance beyond issues of sex. We are at least aware of this shortcoming and would gladly incorporate aspects on, for example, the role of crime in the shaping of male stereotypes in the criminal or forensic psychiatric system within the ongoing projects carried out in collaboration with researchers from the fields of philosophy and legal science.
Henrik Anckarsäter is principal investigator for all the sub-projects described above, in collaboration with, for sub-project 1, Lichtenstein, Långström and Kerekes, for sub-project 2, Lichtenstein, and for sub-project 3, Nilsson. Henrik Anckarsäter assumes scientific responsibility for data collection, analyses, and writing of papers in all the areas described in this project plan, and he is the main tutor or co-tutor for all PhD candidates and other students performing analyses of clinical data and bears the overall responsibility for the funding and carrying out of the analytical work in his research group at the Universities of Gothenburg and Lund, while specific analyses are done within the larger collaborative network (such as twin-analyses by specific softwares in the CATSS). Within the research group headed by Henrik Anckarsäter, Dr Thomas Nilsson is responsible for the follow-up studies and for studies of prediction and victimization; Dr Nóra Kerekes is responsible for the mathematical analyses of epidemiological data and therefore for the epidemiological studies in this application, and MSc Sebastian Lundström is responsible for the longitudinal follow-up studies of the twin cohort. For molecular biology and genetics, collaborations are established with, among others, Professors Kaj Blennow, Henrik Zetterberg, Elias Eriksson, and Thomas Bourgeron.
Henrik Anckarsäter is the recipient as main applicant of large funds from all major Swedish funding sources (National Research Council since 2003, ALF Lund and Gothenburg, FAS, RJ, and the state authorities KV, SiS and RMV), was awarded the prize for most successful younger researcher at the UMAS 2007 in competition with all other medical specialities at the University Hospital and is elected board director in the International Academy of Law and Mental Health (IALMH), the largest and most important international association in forensic psychiatry and related disciplines working with mentally disordered criminal offenders.
In the planning of the budget for this application, the great need in Swedish forensic psychiatry to recruit new clinicians (psychiatrists and psychologists) who are also interested in research has been considered. A detailed work-plan is attached in D Scientific report. The budget for this application includes three post-doctoreal positions, one PhD candidate and one research assistant/PhD candidate. Thomas Nilsson is currently funded from ALF and Nora Kerekes and Sebastian Lundström from the Prison and Probation Services, thus not holding stable, funded positions for the up-coming years, even if they may be able to go on doing research if external funds are obtained at the current level. Should we receive considerably larger funds from the research council in response to this application, these three persons could have stable funding for the up-coming years, and in this case probably leaving room for new recruitments on the grants that currently cover their salaries.
Several new posts in research are also needed for MD residents, psychologists, social workers in training, and nurses doing MSc programmes. The present project holds huge promise to meet this need in Gothenburg and Malmö/Lund. As described above, a PhD thesis is planned for Örjan Falk, a medical student, in sub-project 1, and for Robert Sandin, also a medical student, in sub-project 2. Data are sufficient for several more PhD projects, which will be opened for external recruitment depending on available funding. The ongoing PhD candidates (Larson, Wallinius, Hansson, Lund, Gustavson, Ståhlberg) are all covered by available funds from other sources (mainly ALF). Thomas Nilsson will apply for new PhD posts for sub-study 3 in the upcoming years. As Henrik Anckarsäter has full tenure for research with ALF grants and most data for the three sub-projects has been collected, it may be stated that the project is feasible almost without more funds, but that such a low-budget variant would entail fewer and less methodologically advanced publications and a waste of possibilities for human development in an area where this is most needed.