Studying mental disorders among perpetrators of mass murder–suicide: Methodological challenges and promising avenues for new research

Abstract

Despite tremendous public and media interest in murder–suicide, scientific research that sheds new light on either the prevalence of mental disorders among perpetrators of mass murder–suicide or the role of mental disorders when they appear relevant has been rare. I outline several assumptions and challenges that may encumber scholarship in this area but also identify some promising avenues for future research. Greater scientific understanding of any associations between mental disorders and murder–suicide could lead to more evidence-based interventions that could help prevent these lethal attacks.

In recent decades, incidents of mass murder–suicide—including mass shootings and suicide terrorism—have received a great deal of attention from the public, the media and scholars. However, research that sheds new light on either the prevalence of mental disorders among these perpetrators or the role of mental disorders in their violent behaviour has been rare. This is partly evidenced in the review by Theodorou and colleagues (2024) in this issue of CBMH, which found only 60 qualifying studies on family, mass shooter and terrorist perpetrators combined. Below, I outline several assumptions and challenges that may encumber scholarship in this area and then identify some promising avenues for future research.

Although mental illness is sometimes used as an independent variable in studies of mass murder–suicide—along with other individual characteristics such as age, sex, race, employment status and criminal record—research that is specifically designed to generate better understanding of perpetrators' mental state is rare.

One reason may be the assumption that all perpetrators must have had mental disorders due to the extreme nature of their crimes. If this were true, there would not be much more value in documenting the presence of this variable than in studying how many perpetrators breathed air; in both cases, the results would be self-evident and known in advance. There is a long history in which other acts of mass violence—including genocide and terrorism—were assumed to be the product of mental illness and psychopathology, because most people could not imagine anyone would do such things unless they were ‘crazy’ (Lankford, 2008). However, as the National Council for Mental Wellbeing (2019) has wisely warned, it is not accurate that all ‘acts that seem incomprehensible to the average person are due to mental illness’ (pp. v–vi). Determining which deviant behaviours are commonly associated with mental disorders can only be determined through scientific inquiry.

The opposite assumption—that mental disorders are largely irrelevant to some types of mass murder–suicide (Atran, 2003; McCauley, 2002; Pape, 2005; Skeem & Mulvey, 2020)—may inhibit research as well. For those who adopt this perspective, there would be little point in studying perpetrators in depth. Unfortunately, some media articles that cite scholars have promoted this view, stretching the evidence to claim that ‘Most Mass Shooters Aren't Mentally Ill’ (Rosenwald, 2016), ‘Connecting Mental Illness and Mass Shootings Misses the Point’ (McCausland, 2017) and ‘Mass Shootings are Not a Mental Health Problem’ (Mikkelsen, 2022). Similarly, a prominent professor and psychiatrist has been widely quoted making statements such as ‘There's no mental illness whose symptoms are shooting anyone else’ (McCausland, 2017; Miller, 2021). Although this is technically true—and it is important to recognise that most people with mental disorders are not violent (National Council for Mental Wellbeing, 2019)—the statement is also misleading. Mental disorders can contribute to many behaviours that are not specifically listed in diagnostic manuals as symptoms, and it is possible that the interaction of mental disorders and other factors play an important role in some forms of extreme violence.

In fact, a more reasonable assumption is that most perpetrators of mass murder–suicide may have had a mental disorder, given the clear relationship between suicide and mental illness. Although research has found a ‘modest link’ between mental illness and violence (National Council for Mental Wellbeing, 2019, p. vi), there is a stronger association with suicide (Brådvik, 2018; Joiner et al., 2017; Too et al., 2019; World Health Organization, 2023). In cases where people die by suicide and no mental health problem appeared present, it is sometimes due to missing information (Joiner et al., 2017) and sometimes because the suicidal decision was the product of social or situational forces (e.g. coercion, threat, crisis) or sudden impulse (Lankford, 2016; World Health Organization, 2023). Although family killers may sometimes lose their temper, kill their loved ones and then choose suicide to escape their crisis (Turanovic et al., 2022), public mass shooters and suicide terrorists are usually acting of their own volition and committing premeditated attacks, not impulsive ones (Lankford, 2018; Merari, 2010; Pape, 2005). Accordingly, when they die by suicide or express interest in that fatal outcome, their mental state may be similar to those of suicidal individuals with mental disorders. Researchers who are satisfied by their knowledge of this connection may therefore decide not to investigate the subject for themselves.

For scholars who are interested in studying the mental state of these perpetrators, there are also significant methodological challenges. First and foremost, these individuals differ from most who engage in crime because they are often deceased by the time they are identified by researchers (Lankford, 2018; Peterson & Densley, 2023). This means they can no longer be interviewed or directly assessed, so the best methods for measuring their mental state are no longer available. In some cases, this limitation is compounded because the offenders also killed a spouse, family member, co-worker, fellow student or other acquaintance who knew them well. Family killings are actually the most common form of mass murder (Turanovic et al., 2022), and some mass shooters who attacked in public killed a romantic partner or family member in addition to other victims (Peterson & Densley, 2023). As a result, this tragic loss of life eliminated someone who closely witnessed the offender's pre-attack behaviour and may have been able to provide valuable information.

Because post-attack interviews of these types are often not an option, some researchers instead rely on data on whether perpetrators of mass murder–suicide were formally diagnosed with a mental disorder prior to their crimes. With this approach, researchers can simply report the results that came from clinicians, rather than attempting, albeit retrospectively, to assess perpetrators' mental health for themselves. Unfortunately, such information is often incomplete and yields significant underestimates when compiled in the aggregate. These crimes are disproportionately committed by young men, who are the demographic group most likely to avoid doctors (O’Hara & Caswell, 2013). Even if offenders have come in contact with a clinician who could have assessed them, they may have deliberately concealed symptoms of a mental disorder to avoid shame and stigma, protect their image or avoid being stopped from accomplishing their homicidal or suicidal goals (Lankford, 2016). Accordingly, findings from the United States Secret Service (Vossekuil et al., 2002) and Federal Bureau of Investigation (Silver et al., 2018) suggest that only a small proportion of school shooters, mass shooters and active shooters were diagnosed with a mental disorder before their attacks, but that these diagnoses represented only a fraction of all cases where a mental health problem appeared present.

Another option for researchers who wish to identify mental disorders among perpetrators of mass murder–suicide is to conduct retrospective assessments or psychological autopsies. In many ways, this is similar to the process by which official authorities investigate the antecedents of a crime, but researchers often lack the ability to access confidential information or personally interview people who were close to the perpetrator. Instead, they typically consult primary source documents in which perpetrators shared their thoughts (e.g. manifestos, suicide notes, journal entries, martyrdom videos or online posts), biographical details published in official investigations, media reports with quotes from witnesses who observed things the perpetrator did or said and other information from reliable sources that pertain to the individual's mental state.

Indirect assessments of this type sometimes occur in the legal context, when experts testify about the mental state of defendants who have refused to be evaluated (Kooijmans & Meynen, 2017; Wilhelm, 2022); in the intelligence community, when analysts create psychological dossiers on foreign adversaries from thousands of miles away (Dyson, 2014) and in the media, when consulting experts are asked to comment on individuals they have never met (Holloway, 2003). However, this information is often viewed with scepticism and, at times, appropriately so. While psychological distance between researchers and perpetrators can increase objectivity and reduce the influence of personal connection, it can also be a major limitation. Given the labour involved, the potential for second-guessing and criticism, and the difficulty of ever achieving perfect clarity about a perpetrator's mental state, many researchers avoid this method altogether. Again, this could help explain the relative scarcity of original studies found by Theodorou and colleagues (2024).

Despite these challenges, there are several ways researchers can continue to generate new insights on the prevalence and role of mental disorders in mass murder–suicide. First, it is possible to combine findings from multiple methodologies. For example, mass shooters and suicide terrorists do not always die; sometimes they had suicidal intent but changed their mind at the last minute, or they were arrested before they attacked or before their attacks reached a suicidal conclusion (Peterson & Densley, 2023; Silva, 2022). In these cases, such individuals can potentially be interviewed and have their mental state directly assessed (Lankford, 2018; Merari, 2010). Those findings can be combined with the results from pre-attack diagnoses, retrospective assessments and psychological autopsies of other perpetrators of the same type, thus creating a fuller picture of the association between mental disorders and that form of violence.

Another exciting option is for researchers to study mental disorders that may have gone undetected by testing for correlates that could indicate their presence. For example, Lankford and Cowan (2020) recently compared public mass shooters who had clear signs of a mental disorder with public mass shooters for whom no such evidence had been found. If these were genuinely two distinct groups, correlates of mental illness should have been more common among the mentally ill group. However, quantitative analyses showed no significant differences based on their frequency of being raised by a single parent, being bullied, experiencing childhood trauma, having a parent commit suicide, having a criminal record/arrest record, or dying by suicide or confrontation with law enforcement as a direct result of their attacks (Lankford & Cowan, 2020). This suggests the two groups may have been more similar psychologically than had previously been assumed. In fact, the only statistically significant difference between the two groups was the amount of missing information, which was much higher among mass shooters for whom a mental disorder had not been detected.

Trying to understand the functional role of mental disorders in mass murder–suicide is also a complicated but intriguing path for future research. For instance, Peterson et al. (2022) attempted to analyse how often public mass shooters were known to have experienced psychosis and whether their delusions or hallucinations appeared directly linked to their purported motivations for attacking. They found that in cases where psychosis could be detected, perpetrators were about evenly split between three categories. Some seemed largely driven by their psychosis and cited no other motive for their crime, others were clearly influenced by their psychosis but also had another motive or grievance that played a role, and the remainder seemed to primarily attack because of some adversity in their life, such as being fired from work, with the relationship between their psychosis and their violent actions being far less clear (Peterson et al., 2022). Although measuring the relative influence of different factors on a perpetrator's mental state is fraught with imprecision, it is helpful to learn from such creative attempts to study this subject.

A final promising approach is research that accounts for the timing of different factors in perpetrators' lives. For example, Lankford and Silva (2021) examined some of the deadliest mass shooters since 1999 and found that perpetrators' mental health contacts often began more than a decade before their mass shootings and commonly preceded their work or school problems and acquisition of firearms. However, many perpetrators' final mental health contact was more than a year before their attacks, which suggests that in cases where they had been receiving treatment, that treatment had stopped by the time they committed a mass shooting. Other evidence indicated that some mass shooters seemed to reach a ‘tipping point’ when they became focused on acquiring weapons and preparing for their attacks rather than getting help or improving their lives. Overall, the authors concluded that mental health problems were typically a constant in these perpetrators’ lives, not a new variable that inevitably or automatically caused them to attack, and that lack of treatment and interactions between psychological problems and other life stressors may have contributed to their acts of mass violence (Lankford & Silva, 2021).

Although mental illness is often one of the major discussion points in the aftermath of mass murder–suicide, it is surprisingly understudied due to several common assumptions and methodological challenges. However, significant difficulties and limitations on research in this critically important area should not prevent scholars from delving into it and following promising approaches that could generate new insights. Although understanding the prevalence of mental disorders among mass shooters, suicide terrorists and other perpetrators is an important first step, greater knowledge of the functional role of mental state in extreme violence is also needed. Theodorou et al. (2024) drew some themes from the literature that could also shape future research. In turn, this information could lead to more evidence-based interventions that could help prevent these lethal attacks.