Abstract
Despite rising rates of suicidal thoughts and behaviors in children, little is known about best practices for conducting suicide risk assessments in this population. The acquisition of the concept of death occurs during childhood, and thoughts and preoccupation with death can be developmentally normative. This review highlights a critical gap in knowledge about how children with suicide ideation understand death—a foundational issue in assessing suicide risk. First, we provide an overview of studies that examined the concept of death and related constructs (e.g., preoccupation with death) among children who experienced suicidal ideation or attempted suicide, many of which date back to the 1970s and ’80s. We describe indirect methods of assessing death cognitions, such as the Death Implicit Association Test, representations of death/suicidal themes in play, and the role of exposure to these themes through screens (e.g., social media, videogames). Given the potential impact of sociocultural changes since the 1970s–1980s on how children understand death, gaining a contemporary perspective on how children think about death and its relationship to suicide risk is warranted. Guided by a developmental approach, we provide recommendations for future areas of research to advance this field and inform the design of developmentally appropriate risk assessments and interventions for children.
Plain language summary
The concept of death among children who experience suicide ideation or attempt suicide: The development of the concept of death takes place in childhood and entails the understanding of five dimensions: inevitability, universality, irreversibility or finality of death, cessation, and causality. Historically, some researchers and clinicians have argued that without understanding the concept of death, children could not think about or attempt suicide. This idea has been discarded, yet evidence to contradict this statement is lacking. We conducted a review of studies examining the concept of death and related constructs among children who experience suicide ideation or attempt suicide. Most studies on this topic were conducted in the 1970s and 1980s and suggested that children who think about or attempt suicide may have a distorted concept of death, often making more references to the afterlife and resurrection than children without these behaviors. They tend to describe death as more pleasant and less permanent, report a greater preoccupation with death, and endorse more violent ways of dying. Societal changes since these studies were conducted may have influenced how children now conceptualize death. Gaining an updated perspective on how children think about death and its relationship to suicide risk can help clinicians assess the severity of suicidal thoughts and behaviors and make recommendations for care. Guided by a developmental approach, we provide a critical review of prior studies, discuss alternative ways to assess death constructs (e.g., implicit identification with death, representations of death and suicidal themes in play), consider exposure to death-related themes through screens (e.g., social media, videogames), and propose new research directions to advance the field. We believe these contributions will aid in designing developmentally appropriate risk assessments and interventions for children.
Keywords
Introduction
Historically, suicidal thoughts and behaviors in childhood (5–12 years old) have received minimal attention from researchers, likely due to the sensitivity of the topic alongside skepticism about whether children can think about or attempt suicide. Early publications from the 1970s–1990s established that children can indeed experience suicidal ideation (SI) and engage in suicidal behaviors (Gould et al., 1998; Klimes-Dougan et al., 1999; Pfeffer et al., 1979, 1980). However, interest in this topic seemingly waned until two studies published in 2015 and 2018 highlighted a concerning increase in suicide deaths in the past 20 years, especially among Black children (5–11/12 years) (Bridge et al., 2015, 2018). Subsequent research reported increasing numbers of emergency department (ED) visits for SI and suicide attempts (SAs) in 5-to-11-year-old children (Burstein et al., 2019). These and subsequent publications documenting the prevalence of suicidal thoughts and behaviors in childhood (DeVille et al., 2020; Lanzillo et al., 2019; Ortin-Peralta et al., 2023; Whalen et al., 2015) renewed the interest in this topic and raised fundamental questions concerning the developmental precursors of SI and SA in childhood (Oppenheimer et al., 2022). However, despite these developments, there remains a striking absence of evidence-based recommendations to guide risk assessments and interventions for children experiencing SI/SA (Ayer et al., 2020; Cwik et al., 2020).
A gap in the literature that may contribute to the limited attention placed on childhood suicide is the lack of understanding regarding the meaning of death and suicidal thoughts in childhood (i.e., do children really mean it when they say they want to die?) (Scheeringa, 2016). Although some researchers and clinicians have previously argued that without mastering the concept of death, children cannot think about or attempt suicide (Pfeffer, 1986), this idea has now been largely discarded, yet evidence to contradict this statement is lacking. As such, the value of inquiring about children’s concept of death in the context of suicide risk assessments and as an intervention target remains unexplored. This gap has direct clinical implications, as clinicians may have difficulties determining the level of suicide risk when children directly or indirectly express thoughts or preoccupation with death or dying during risk assessments, or when such thoughts are reported by caregivers. Identifying what types of death thoughts and conceptualizations are common among children who are thinking about suicide and the relationship of such thoughts with the severity of SI/SA has the potential to improve suicide risk assessment in children.
The goal of this narrative review is to provide a broad overview of the limited literature examining the death concept and related constructs (e.g., perception of death as a pleasant state) among children who have experienced SI/SA. Additionally, we discuss indirect ways to assess death cognitions, such as the Death Implicit Association Test (Death-IAT), representations of death and suicidal themes in play, and exposure to these themes through screens (e.g., social media, videogames). This review aims to synthesize current knowledge on how children understand death in relation to suicide risk, offering a foundation to ground new lines of research and inform the development of age-appropriate risk assessments and interventions.
Development of the concepts of death and suicide
The acquisition of the biological concept of death is a developmental milestone in childhood. Current conceptualizations of death support a progression throughout five dimensions: inevitability (living things die eventually), universality or applicability (death happens to all living things), irreversibility or finality of death (the dead cannot come back to life), cessation or non-functionality (death is characterized by bodily processes ceasing to function), and causality (death is ultimately caused by a breakdown of bodily functions) (Slaughter & Griffiths, 2007).
Studies across different countries show that around the age of 4–5 years children begin to grasp the dimensions of inevitability and irreversibility, followed by universality and cessation (first of biological and then psychological functions), and finally causality (Panagiotaki et al., 2015, 2018; Slaughter & Griffiths, 2007). By ages 8–10 years, most children have acquired the biological concept of death.
The first youth study, to our knowledge, that examined the concept of suicide included children recruited from a psychiatric inpatient unit and healthy children from a local school (Carlson et al., 1987). When children were asked about causes of death, older children (11–13 years) reported suicide more often than younger children (8–10 years) in the inpatient group. A similar percentage of inpatient (77%) and healthy (76%) children knew or had heard of someone who had attempted or died by suicide, with television or the news being the main source of knowledge. When asked means of suicide, children frequently described shooting and stabbing, followed by jumping or overdosing. Feeling crazy, having a hard time, and self-hatred were listed as reasons for wanting to kill oneself (Carlson et al., 1987).
In the 1990s, Mishara conducted two studies in which children (grades 1-5) were asked about the concepts of death and suicide (Mishara, 1999; Normand & Mishara, 1992). The researchers observed that the wording of the suicide-specific questions influenced children’s responses. At first, children were asked if they knew the meaning of the word “suicide.” If they answered “no”, the questions concerning suicide were skipped (Normand & Mishara, 1992). In the second study, if children answered “no” to the first question, the word “suicide” was replaced with “killing oneself” and the question was re-asked. This change in wording resulted in more “yes” responses and children then received additional questions about their understanding of suicide (i.e., intentionality, motives) (Mishara, 1999). Both studies showed that children’s understanding of death and suicide increased with age. Interestingly, in analyses with the full sample, understanding of suicide was not significantly associated with understanding of death. However, when the sample was limited to children who possessed some knowledge of “suicide” or “killing oneself”, the concepts of death and suicide were positively associated (Normand & Mishara, 1992). These findings suggest that children start to develop the concept of death prior to that of suicide, and they might integrate the concept of suicide as one of many potential causes of death.
When children are developing the concept of death, thoughts and preoccupation with death are frequent, typically reflecting curiosity or personal experiences, and not necessarily signaling psychopathology (Stoep et al., 2009). To inform risk assessment, clarifying whether the conceptualizations of death among children who are thinking about suicide are more frequent and/or qualitatively different (i.e., more or less advanced) than among children who have never experienced SI/SA is warranted (Oppenheimer et al., 2022; Scheeringa, 2016).
Concept of death and death-related constructs among children with suicidal thoughts
Summary of the Main Findings From Studies Examining the Concept of Death and Related Constructs in Children Who Have Experienced Suicide Ideation or Attempted Suicide.
Boxes with information summarize the findings for children with SI/SA compared to children without SI/SA. An empty box indicates that a death dimension or construct was not examined in the study. All studies were conducted in the USA, except for the two studies by Orbach & Glaubman, which were conducted in Israel. D = depressed; SI = suicide ideation; SA = suicide attempts. a, b, and c refer to three samples of children recruited from (a) a psychiatric outpatient clinic, (b) an inpatient service; and (c) schools.
Around the same period, Pfeffer and colleagues conducted a series of studies on death attitudes in samples of children, 6–12 years, from inpatient and outpatient settings, alongside matched healthy controls (Pfeffer et al., 1979, 1980, 1982, 1984, 1986). The authors developed the Child Suicide Potential Scales, a battery of eight scales that included the Spectrum of Suicidal Behavior scale (non-suicidal, SI, suicidal threat, mild SA, serious SA), and the Concept of Death scale, which assessed three attitudes toward death: preoccupation with death (e.g., thoughts of their own death, worries about family member dying), perception of death as pleasant (whether a person goes to a better or worse place after death), and finality of death (whether they think of death as final or temporary) (Pfeffer et al., 1979). Across studies, children with SI/SA were more preoccupied with death than “non-suicidal” children (Pfeffer et al., 1979, 1980, 1984, 1986). Findings about differences in finality and pleasant state were less consistent (Pfeffer et al., 1979, 1980). Pfeffer and colleagues argued that the children’s emotional state when thinking about suicide may lead to the distortion of death perceptions and the consideration of death as a “means of eliminating stress and attaining peace” (Pfeffer et al., 1980, p. 709).
This topic was not examined again until recently, when Hennefield and colleagues conducted a study with 139 3-7-year-olds with and without recent SI (Hennefield et al., 2019). Children completed the Death Interview (Slaughter & Griffiths, 2007), a structured interview that inquiries about the five dimensions of the concept of death. Contrary to the distortion identified in ‘70s/’80s studies, depressed children with SI scored higher on the Death Interview (i.e., more advanced understanding of death) than depressed children without SI and healthy children. The three groups did not differ on any of the individual dimensions except for causality. Specifically, depressed children with SI were more likely to name a violent event as a cause of death than the comparison groups (Hennefield et al., 2019).
There are several hypotheses that may explain the different results between Hennefield et al. (2019) and prior studies. Hennefield’s study included younger children, many of whom still had an immature concept of death, which may have precluded identifying differences in the individual dimensions. Hennefield et al. (2019) also relied on caregiver report of child SI, whereas earlier studies relied on child report (Pfeffer et al., 1980). Low caregiver-child agreement on youth SI/SA has been documented, thus using caregiver reports likely limited the sample to children who had verbalized suicidal thoughts or acted upon them (DeVille et al., 2020). Nonetheless, a second study conducted by the same research team showed that, based on caregiver report, preoccupations with death were more common among young children with SI/SAs than among children with non-suicidal self-injury (NSSI) or no SI/SA/NSSI (Luby et al., 2019), aligning with Pfeffer’s findings and supporting the importance of including multiple informants when assessing SI/SA.
Indirect measures of death-related constructs
Most studies on the concept of death and related cognitions in children with SI/SA have relied on self-report measures. Indirect measures may be particularly relevant as children may have limited recollection, introspective abilities, and language to describe internal states, especially in circumstances of elevated emotional arousal and mental health concerns (Bourchier & Davis, 2002; Jenkins & Greenbaum, 1999; Sendzik et al., 2017). Examples of indirect measures that have been used with youth with SI/SA include child representations of death and suicidal themes in play, and the implicit identification with death.
Three recent articles have examined death/suicidal themes in play in children, ages 3–6/7 years, recruited for studies conducted by the same research group (Hennefield et al., 2022; Luby et al., 2019; Whalen et al., 2015). In two studies, caregivers completed structured interviews to report on child’s death/suicidal themes in play and drawings (e.g. “Has s/he ever engaged in fantasy play that persistently involves death or dying?”), and lifetime SI, SAs, and NSSI (Luby et al., 2019; Whalen et al., 2015). Whalen et al. (2015) found that children with death/suicidal themes in play but no SI differed neither demographically nor clinically from children with SI but no death/suicidal themes in play. However, children with only death/suicidal themes in play were more likely to be boys and meet criteria for depression, ADHD, and oppositional defiant/conduct disorders than healthy children (Whalen et al., 2015). Luby et al. (2019) found that young children with SI/SAs expressed death/suicidal themes in play and drawings significantly more often than children without SI/SA/NSSI (Luby et al., 2019). In the third study, authors used Story Stem Narratives in which young children were provided narratives to develop an ending with doll figures to capture interpersonal dynamics and conflict resolution tendencies (Hennefield et al., 2022). Children’s responses were coded into four themes: Accidental Harm/Death, Violence/Homicide, Suicidal Acts/Ideation, and NSSI. Caregivers reported on children’s SI. Compared to children with no SI, those with lifetime active SI were 3.59 times more likely to display violence/homicide endings. The narrative endings of 28.5% of children with current active SI received a code of Suicidal Acts/Ideation versus 3.4% of the narratives of the children without current active SI (Hennefield et al., 2022). Altogether, these findings suggest that the manifestation of death and suicidal themes in play/drawings could signal high suicide risk for younger children.
Implicit associations occur outside the conscious awareness and are less subject to alteration due to unwillingness or inability to report (Banse et al., 2001). The Death Implicit Association Test (Death-IAT) is a computerized task that measures an individual’s reaction time when classifying words related to death and life with “me” or “not me” (Nock & Banaji, 2007). Studies have found that implicit identification with death is positively associated with presence, severity, and duration of SI, especially when adolescents are assessed close to the suicidal event (Brent et al., 2023; Glenn et al., 2019; Toukhy et al., 2023). In a recent study, 10-to-17-year-olds admitted to an ED for SI or SA reported an improvement in mood and a decrease in desire to die after completing the task (Shin et al., 2023), providing initial evidence on the safety of the Death-IAT. Other studies have successfully administered IATs on differ topics (e.g., alcohol, race) to school-age children (Noel & Thomson, 2012; Steele & Lipman, 2023), but no studies to date have used the IAT to examine implicit identification with death in children under age 10.
Societal changes and cultural diversity
Most findings in children, although informative, are dated, and need to be replicated. Gaining an updated perspective of how children think about death when they are experiencing SI/SA can help clinicians incorporate this information when assessing severity of SI/SA and making recommendations for care.
Societal changes since the first studies were conducted, may have influenced how children grasp the death constructs. In Western societies, for example, greater geographical mobility and changes in the family composition have reduced children’s contact with extended family and exposure to mourning rituals (Longbottom & Slaughter, 2018). Although discussions about death with parents has been linked to a more advanced concept of death (Hunter & Smith, 2008; Yang & Chen, 2006), many Western parents feel uncomfortable having these conversations and might avoid them to protect their children from experiencing negative feelings (McGovern & Barry, 2000; Rosengren et al., 2014). On the other hand, compared to children in the 1970s/‘80s, children today are exposed to greater levels of death, violence, and suicidal content through screens, where those concepts are often portrayed inaccurately (Bridgewater et al., 2021; Tenzek & Nickels, 2019). In the television series “13 Reasons Why”, the main character narrates her story from the afterlife after having died by suicide, providing an example of an inaccurate depiction of irreversibility and cessation. Little research has addressed how societal changes and increased exposure to screens and social media have impacted children’s acquisition of the concept of death and the development of suicidal thoughts and behaviors. These societal changes regarding how, when, and what aspects of death children are exposed to, may also account for different findings between the studies conducted in the ‘70s/’80s and Hennefield et al. (2019)’s study.
Regarding suicide risk, a recent study found that more screen time (i.e., watching TV shows/movies, streaming videos, texting, playing video games, video chatting, and social media) was associated with lifetime SI/SA assessed two years later (Chu et al., 2023). The content of screens and social media may also be relevant for suicide risk, as “13 Reasons Why” demonstrated. The release of the show was followed by an increase in SI, self-harm, and suicide in adolescents (Bridge et al., 2018; Cooper et al., 2018; Niederkrotenthaler et al., 2019; Sinyor et al., 2022) and to some extent children (Le et al., 2022). The detailed portrayal of a suicide and the inaccurate characterization of the protagonist after her death could be potential reasons why increases in suicidal thoughts and behaviors/related behaviors (i.e., self-harm) occurred.
Limitations and future studies
One key limitation of the 70s/’80s studies was the time frame used for the assessment of SI and SAs (i.e., either lifetime or unspecified), which has contributed to the lack of clarity concerning the timing of the thoughts of suicide, distortion of the concept of death, and which may come first. Determining the clinical value of the concept of death and related constructs requires a more refined timing and content of the assessments of the death dimensions and related constructs (i.e., during or close in time to when children are thinking about suicide). Additional limitations include cross-sectional designs, small samples, lack of diversity, and the utilization of self-report measures from one informant (either child or caregiver). As such, longitudinal studies that rigorously combine reports from multiple informants are needed to identify the best ways to inquire about the death concept and related constructs and determine their value to assess and predict suicide risk among children. Studies involving multiple informants would be preferable not only to gather multiple points of view on the same experience but also to shed light on how clinicians can integrate information collected from multiple informants into the assessment of mental health problems (Cloutier et al., 2010).
Comparisons of children from diverse cultural backgrounds may be useful to examine how different traditions, beliefs, and other cultural-specific factors influence the understanding of death and desire to die. For example, being raised in a religious family or context, especially when the religious beliefs include the afterlife, seems to affect the development of the biological concept of death (Florian & Kravetz, 1985; McIntire et al., 1972). Children raised in religious environments are more likely to report non-natural reasons for death such as “Selection of God” and have a less developed understanding of cessation and irreversibility, believing that some psychological functions continue after death compared to those raised in non-religious contexts (Bering et al., 2005; Panagiotaki et al., 2018; Yang & Chen, 2006). Noteworthy, the way children manage biological or spiritual conceptualizations of death changes with age. Most children hold either biological or spiritual conceptualizations of death until adolescence, when they become capable of holding co-existing conceptualizations of death, biological and spiritual/metaphysical, without tension (Legare et al., 2012; Panagiotaki et al., 2018). This duality emerges once children have acquired the biological concept of death, are able to think abstractly, and have accumulated extensive socio-cultural experiences to inform their knowledge about death (Legare et al., 2012). The interplay between children’s incapability of holding co-existing conceptualizations of death along with their exposure to afterlife beliefs could lead to an increase or decrease of suicide risk.
Novel additions would be the consideration of children’s understanding of the concept of suicide, as well as the inclusion of indirect measures of death/suicidal themes. The use of the Death-IAT could help gain an unbiased report of how death is implicitly associated with ‘self’ among children experiencing SI/SA. While there has not yet been research specifically on the Death-IAT with children, studies involving other types of IATs, suggest that children can reliably complete these tasks. For example, Williams et al. (2016) found no significant differences in the reliability of the Race-IAT when comparing children and adults. However, although this methodology holds promise for tapping into children’s implicit conceptions of death, future studies should assess the validity and reliability of the Death-IAT in children to ensure that it accurately captures implicit associations rather than reflecting developmental differences in cognitive processing or understanding of the task itself.
Furthermore, expanding the assessment of death/suicidal themes in play and drawings by inquiring about frequency and exposure to death and suicidal content through screens may be particularly relevant for school-age children. These studies could provide clarity concerning whether specific frequency or content on the screens is associated with the development of the concept of death and SI in childhood. Another potential way to evaluate the concept of death in childhood is through the use of drawings (Bonoti et al., 2013; Vázquez-Sánchez et al., 2019). However, this technique has not yet been applied to children experiencing SI/SA. Visual methods could offer a developmentally appropriate alternative for assessing children’s understanding of death that do not rely solely on verbal report.
Building on Orbach and Glaubman’s argument about children’s distortion of the death dimensions as a defense mechanism to overcome their own fears about dying (Orbach & Glaubman, 1979), a related construct that can be incorporated is fear of death (Lester, 1967). In support of this addition are recent findings among adolescents receiving intensive outpatient care for active SI or recent SA. Fearlessness of death was positively associated with the level of planning and lethality of the adolescents’ SAs (Krantz et al., 2022) and predicted higher risk for attempting suicide and higher lethality of a SA at follow-up (Ferm et al., 2020; Krantz et al., 2022).
A challenge researchers may encounter when conducting studies on this topic may be the reluctance of caregivers to agree to have their children answer questions about their SI and death conceptualizations. There is empirical evidence that asking adolescents (aged 12+) these questions do not increase suicide risk, and it may even lead to small reductions in SI and lower likelihood of attempting suicide (Blades et al., 2018). Although this work has not yet been conducted with children, we are not aware of any compelling arguments for why risk would be different from adolescents. Reminding caregivers about the low impact that asking these questions has on suicide risk and providing resources to caregivers on how to talk about these topics with children under 12 years could help ease their worries (University of Utah Health, 2022; Ackerman, 2022).
Conclusions and clinical implications
Suicide rates in childhood have increased in the past decades, however, little is known about best practices to guide risk assessments (Ayer et al., 2020). Gaining an updated perspective of how children think about death and how this relates to suicidal thoughts and behaviors can help clinicians inquire about death constructs when assessing suicide risk and incorporate this information into treatment plans. Findings from future studies could inform clinicians on how to assess children’s concept of death and related constructs to determine severity and predict suicide risk trajectories. If a unique way of thinking and comprehending death emerges among children experiencing SI/SA, these cognitions could also serve as targets for intervention to decrease suicide risk in children.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute of Mental Health LH, (K01MH127412); AHS, (R01MH133226), (P50MH127476 sub-project 7957), (R01MH125905) and the American Foundation for Suicide Prevention AHS, (YIG-1-151-19).
Author biographies