Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 1.
Published in final edited form as: Child Adolesc Ment Health. 2021 Mar 28;26(4):331–338. doi: 10.1111/camh.12454

Characterizing Children Hospitalized for Suicide-Related Thoughts and Behaviors

Marisa E Marraccini 1, Christopher W Drapeau 2, Rachel Stein 3, Cari Pittleman 1, Emily N Toole 1, Molly Kolstad 1, Amanda C Tow 4, Shannon M Suldo 5
PMCID: PMC8476654  NIHMSID: NIHMS1718193  PMID: 33779031

Abstract

Background:

Despite alarming increases in suicide deaths among pre-adolescent children, knowledge of the precipitants of suicide risk and the characteristics of children who seek treatment for suicidality is limited. This study’s purpose is to describe children (ages 6–12) hospitalized for suicide-related concerns and compare demographic and diagnostic differences between children and adolescent (ages 13–18) patients.

Methods:

This retrospective study analyzed medical records of 502 children and adolescents ages 6–18 admitted for suicide-related risk to one psychiatric inpatient hospital in southeastern United States between 2015–2018.

Results:

Patients were predominantly White (63.5%), female (64.5%), and non-Hispanic/Latinx (85.1%). We conducted descriptive analyses and a series of logistic regressions comparing children and adolescents with data extracted from discharge summaries, (i.e., primary reasons for admission, environmental stressors, and diagnostic categories). Common environmental stressors included school (63.2%) and family (60.7%), and the most common diagnosis included depressive disorders. Compared to adolescents, children were more likely to be Black (OR=1.99), male (OR=1.94), and receive neurodevelopmental disorder (aOR=3.0) or trauma and stress related disorder (aOR= 2.6) diagnoses, but less likely to be diagnosed with a depressive disorder (aOR=0.4). Across both age groups, Black patients were more likely to be diagnosed with neurodevelopmental disorders and less likely to receive internalizing disorder diagnoses.

Conclusions:

Characteristics of children hospitalized for suicide-related risk are relatively similar to children dying by suicide, with increased prevalence among children who are Black, male, and identified has having a neurodevelopmental disorder diagnosis. Proactively identifying and providing strengths-based supports for Black boys and families appears critical for suicide prevention in children.

Keywords: Child, Psychiatric Hospitalization, Suicide, Adolescent, Suicide-related thoughts and behaviors

Characterizing Children Hospitalized for Suicide-Related Thoughts and Behaviors

Evidence documenting that pre-adolescent children younger than 13 engage in suicide-related thoughts and behaviors (STB), including suicidal ideation, suicide plans, and suicide attempts, has been accumulating over the past few decades (Miller, 2019; Tishler et al., 2007). There was a 276% increase in the suicide rate for pre-adolescent youth (herein referred to as children) ages 6–12 from 2008–2018 (Centers for Disease Control and Prevention, 2020). The rate at which children are hospitalized for suicide attempts significantly increased between 2008–2015 (Plemmons et al., 2018), with caregiver reports suggesting between 2.4% (Martin et al., 2016) and 3.5% (Luby et al., 2019) of children ages 3–7 attempting suicide.

Although many characteristics of children dying by suicide are comparable to adolescents (e.g., being male, affective disorder), children dying by suicide appear more likely to be Black and have an Attention-Deficit/Hyperactivity Disorder (ADHD) diagnosis compared to older youth (Bridge et al., 2015; Sheftall et al., 2016). Findings from the National Violent Death Reporting System for the years 2003–2012 (Sheftall et al., 2016) indicate that (compared to adolescents) children ages 5–11 who died by suicide were more likely to be Black, die by hanging/strangulation/suffocation, die at home, and experience relationship problems; but were less likely to experience depression or dysthymia. Note that because research focused on suicide in children is an emerging area of inquiry, identified risk factors for suicide may vary based on a range of factors including sample characteristics and study methods.

More generally, psychiatric factors, environmental stressors, and behavioral issues are proposed to intersect to influence STB among children, with children having fewer precipitants than older adolescents (Dervic et al., 2008; Tishler et al., 2007). As described by Dervic and colleagues (2008), preliminary studies do indicate that children attempting suicide are at greater risk for making additional attempts (Brent et al., 1999). Additional risk factors include family psychiatric history, sleep disturbances, school environment, and family conflict (Dervic et al., 2008; Lin et al., 2014; Martin et al., 2016; Mayes et al., 2014; Ridge Anderson et al., 2016; Sarkar et al., 2010; Tishler et al., 2007). Studies exploring STB in clinical samples of children have identified the presence of psychopathology, including symptoms of depression, Post-Traumatic Stress Disorder (PTSD), Oppositional Defiant Disorder (ODD), conduct disorder, and ADHD, as risk factors for suicide, with mixed findings concerning differences between boys and girls (Martin et al., 2016; Whalen et al., 2015; Wyman et al., 2009).

In one of the only studies to compare differences between children and adolescents hospitalized for STB, Sarkar and colleagues (Sarkar et al., 2010) examined data collected from a children’s hospital in Dublin from 2002–2008. Of the 401 youth receiving treatment for STB, 21.9% were under the age of 12, with the remainder between 12–16. Children were less likely to have self-harmed than adolescents, and more likely to be male, have a history of bullying victimization, and a family history of depression.

Variability in precipitants based on age may in part be explained by developmental considerations unique to children, such as the ways STB are expressed, the intentions of children expressing STB, and children’s understanding of death (Luby et al., 2019; Tishler et al., 2007). Children presenting with suicidal concerns may be simultaneously learning about death, making it critical to differentiate natural curiosity about death from thoughts of suicide (Scheeringa, 2016). Children’s cognitive capacity to understand the finality of death and the concept of suicide has been proposed as an important factor for understanding suicide risk. Alongside other cognitive risk factors (e.g., concrete thinking), immature views of death are hypothesized to increase suicide risk (Barrio, 2007). Underdeveloped abstract thinking is proposed to limit existential thoughts about the consequences of death and suicide (Tishler et al., 2007), increasing risk for attempting suicide (Barrio, 2007). Preliminary research exploring views of death and suicide amongst primary school children, however, suggests that even children in first grade understand the finality of death and the meaning of killing oneself (Mishara, 1999).

Irrespective of children’s capacity to understand the lethality of death, mortality data and research show that STB in children can lead to death (Pfeffer et al., 1980). Risk for suicide attempt appears to increase for children with a history of psychiatric hospitalization and having made a previous suicide attempt (Pfeffer et al., 1993), emphasizing the need to better understand the variables and contexts leading to hospitalization (Ayer et al., 2020). Therefore, the purpose of the present study was to describe pre-adolescent children hospitalized with STB and explore differences between hospitalized children and adolescents.

The primary aim of this study was to characterize hospitalized children ages 6–12 with a billing code related to suicide at one psychiatric inpatient hospital in the southeastern United States (U.S.) between the years of 2015–2018. We identified and explored relationships between common reasons for hospitalization, environmental stressors related to suicide risk, and primary diagnoses. We also explored whether demographic characteristics (sex, race, and age) and common reasons for hospitalization predicted the diagnostic categories that children received at discharge. Hospitalizations among children appear more common among boys (Sarkar et al., 2010) and STB has been shown to associate with behavioral disorders such as ADHD and ODD (Whalen et al., 2015). Therefore, we hypothesized that being Black, male, and younger would be (a) positively associated with diagnoses related to behavior problems (e.g., neurodevelopmental disorders including ADHD); and (b) negatively associated with diagnoses related to internalizing symptoms (e.g., depressive and anxiety disorders).

The second aim of the present study was to examine differences between children (ages 6–12) and adolescents (ages 13–18) hospitalized for STB. Compared to adolescents, children who die by suicide are more likely to be Black (Bridge et al., 2015), male and have a behavioral disorder than older children (Bridge et al., 2015; Sheftall et al., 2016). Younger patients also appear more likely to be male compared to older ones (Tossone et al., 2014) and PTSD is associated with STB in children (Martin et al., 2016). Therefore, we hypothesized that compared to adolescents, children would be (a) more likely to be Black and male, and (b) more likely to receive a diagnosis related to problem behaviors and trauma, but less likely to receive diagnoses related to internalizing problems.

There are distinct developmental differences between children and adolescents regarding the expression and intentions of STB and their cognitive capacities for understanding death (Luby et al., 2019; Tishler et al., 2007). Accordingly, we focused on differences between children (ages 6–12) and adolescents (ages 13–18), but also conducted sensitivity analyses to account for the variability within individual development (i.e., the age at which children transition into adolescence varies from child to child; e.g., Euling et al., 2008).

Method

This retrospective study of archival medical data was reviewed and approved by the Institutional Review Board. We obtained medical records of 502 children and adolescents admitted to a large university teaching hospital with a suicide-related billing code (based on suicide-related diagnoses in the International Classification of Diseases, Tenth Revision [ICD-10]) between 2015–2018. Records were limited to children or adolescents aged 4–18 years (the youngest age of children with data were 6 years) with suicide-related billing codes hospitalized for at least 24 hours. Demographic characteristics and information from discharge summaries were extracted from medical charts.

Cases of Patients

Data from 502 unique cases of children and adolescents were included in the study. Duplicate visits for 29 individuals were removed. The sample included children ages 6–12 (n=121) and adolescents ages 13–18 (n=381). Across age groups, children and adolescents had an average stay of 14.0 days (SD=14.0), were predominantly White (63.5%), female (64.5%), non-Hispanic/Latinx (85.1%), and spoke English as their primary language (92.6%; see Supplementary Table S1).

Both children and adolescents hospitalized for suicide risk were included in analyses focused on demographic (age, race, ethnicity, and sex) and diagnostic information. Only children ages 6–12 years were included in analyses related to reason for hospitalization, diagnosis, and stressors. Patient ages of children were six (2.5%), seven (2.5%), eight (7.4%), nine (12.4%), ten (13.2%), eleven (23.1%), and twelve (38.8%) years old.

Data Coding

Demographic information, including age, sex (male or female), race (American Indian/Alaska Native, Asian, Black/African American, White/Caucasian, or other), and ethnicity (Hispanic/Latinx or non-Hispanic/non-Latinx) was extracted from patient records. With the exception of age, all demographic variables included in analyses were coded as binary, with race, ethnicity, and sex dummy coded (i.e., Black or other; Hispanic/Latinx or non-Hispanic/non-Latinx; female or male). We specifically coded race as Black or “other” to explore our study hypotheses related to the prevalence of Black children hospitalized for STB. A variable representing age group (child or adolescent) was also coded.

Primary reasons for hospitalization, diagnoses, and environmental stressors were extracted from clinical notes in patient discharge summaries. Discharge summaries were written by the medical provider caring for the patient at the time of discharge and include hospital course and day of discharge evaluation. Although the dataset was comprised of cases with suicide-related billing codes, specific reasons for hospitalization included a range of behaviors such as suicidal ideation, suicide attempts, and psychiatric symptoms. For example, a child may have been primarily hospitalized for aggressive behaviors but received a suicide-related billing code given additional concerns regarding suicide risk. Similarly, a child who made a suicide attempt and also struggled with suicidal urges would receive a suicide-related billing code and may have both suicidal ideation and suicide attempt coded as the primary reasons for hospitalization in their discharge summaries. Each reason for hospitalization was coded with a binary code (yes/no) to allow for multiple reasons.

Physicians identified multiple stressors contributing to presentation and recorded these in discharge summaries. Because stressors captured a range of issues, they were coded inductively by the first author and emergent themes were identified. The focus was on environmental stressors (e.g., parents, school, and relationships), which encompassed the majority of themes. Stressors related to psychiatric symptoms or psychological functioning (e.g., poor coping skills or command auditory hallucinations) were excluded considering they were also described as reasons for hospitalization and captured in diagnostic summaries. Examples of themes, described in the results section, included family, school, other social/interpersonal difficulties, trauma/abuse, divorce, and recent move. Variables were derived from a total of 18 themes and given a binary code (yes/no).

Primary diagnoses were extracted from discharge summaries, which included one or multiple primary diagnoses. Diagnoses were then coded based on broad categories outlined in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V; Diagnostic, 2013) shown in Supplementary Table S2. Each patient could have multiple primary diagnoses; thus, diagnosis variables were derived from DSM-V categories with binary (yes/no) codes. Note that all concerns for diagnoses, including rule outs and traits (e.g., of personality disorders), were coded as “yes” for each patient.

Data Analysis

Data were analyzed in SPSS Version 26 (IBM Corp., 2019) and MPlus Version 8 (Muthén & Muthén, 2017). To explore our first aim, to characterize children hospitalized for suicide risk, descriptive statistics (e.g., frequency, means) were calculated to provide an overview of patient characteristics. Phi coefficients were calculated to examine the association between binary variables (i.e., demographics, environmental stressors, diagnoses, and reasons for hospitalization). Point-biserial correlations were conducted to examine associations between age and the previously identified binary variables. A series of logistic regressions examined key demographic variables (age, race and sex) and the most commonly reported reasons for hospitalization as predictors of diagnostic categories (depressive, anxiety, trauma and stressor-related, and neurodevelopmental disorders).

To explore our second aim, to examine demographic and diagnostic differences between children (ages 6–12) and adolescents (ages 13–18), we conducted a series of logistic regressions. To capture developmental differences between children and adolescents, age was dummy coded by age group (i.e., age was not expected to linearly predict outcomes, but instead the distinct age periods were expected to associate with outcomes). First, two separate models explored age group (children vs. adolescents) as a predictor of sex (male vs. female) and race (Black vs. other). Second, four separate models were conducted to examine age group as a predictor of youth receiving diagnoses of depressive, anxiety, neurodevelopmental, and trauma and stressor-related disorders, controlling for sex and race.

Because there is no clearly defined age at which children transition into adolescence (Euling et al., 2008), we also conducted the same models described above, excluding children 12 years of age, either focusing on children ages 6–11 or comparing children ages 6–11 to adolescents ages 13–18. Findings from these sensitivity analyses are presented in Supplementary Materials; note, however, that due to power limitations these results should be interpreted cautiously.

Results

We first present findings for a subset of hospitalized children in the 6–12 age range (n=121) and next present findings for all hospitalized youth with suicide-related billing codes (n=502), exploring differences between children (n=121) and adolescents (n=381).

Characteristics of Pre-Adolescent Children Ages 6–12

All cases including children ages 6–12 (n=121) had descriptive reasons for hospitalization in the extracted discharge summaries (see Supplementary Table S2). The primary reasons for hospitalization described in discharge summaries included suicidal ideation (71.9%); depression, anxiety, or trauma (34.7%); aggression or behavioral dysregulation (26.4%); suicide plans (20.7%); and self-injury (16.5%).

Environmental stressors were documented for 117 of 121 patients. The most commonly emerging themes included school (e.g., academic challenges, bullying; 63.2%), family (e.g., family discord, new baby, parental medical concern; 60.7%), and social or interpersonal conflict (e.g., strained friendships; 23.1%).

Primary diagnoses were recorded in discharge summaries for 119 of 121 patients. The most common diagnostic categories for hospitalized children with a suicide-related billing code included depressive (64.7%), trauma and stress related (31.9%), neurodevelopmental (28.6%), and anxiety (18.5%) disorders. Note that neurodevelopmental disorders primarily included ADHD diagnoses (76%), but also included autism spectrum disorder (23.5%) and intellectual disability or borderline intellectual disability disorder (8.8%) diagnoses. Multiple diagnoses were provided for each patient and number of diagnostic categories ranged from 0 to 4 (M=1.60; SD=0.78).

Correlations between demographic variables and common reasons for hospitalization, environmental stressors, and diagnoses provided for children (ages 6–12) are shown in Supplementary Table S3. Male sex and younger age were significantly and negatively related to the environmental stressor of interpersonal conflict, as well as multiple reasons for hospitalization and diagnoses. Black race was significantly and positively related to having a family stressor and receiving a neurodevelopmental disorder diagnosis, and negatively associated with receiving an anxiety disorder diagnosis. Hospitalization for aggression or behavioral dysregulation was significantly and positively associated with neurodevelopmental disorder diagnosis, and also negatively associated with depressive disorder diagnosis. Hospitalization for anxiety, depression, or trauma showed the reverse pattern, with a positive association with a depressive disorder diagnosis and negative association with a neurodevelopmental disorder diagnosis.

Correlates of Diagnoses within Pre-adolescent Children Ages 6–12

To test our hypothesis that being male and younger would be associated with receiving diagnoses related to behavioral problems (neurodevelopmental disorders) and being female and older would be associated with receiving internalizing diagnoses (depressive disorders, anxiety disorders) in children, we examined sex, age, race and common reasons for hospitalization (suicidal ideation; aggression or behavioral dysregulation; and depression, anxiety or trauma) as predictors of diagnoses (see Table 1). No variables significantly predicted children receiving a trauma and stress related disorder diagnosis. Age was a significant predictor of receiving a depressive disorder diagnosis, with older age related to receiving the diagnosis. Although neither sex nor age were significant predictors of receiving a neurodevelopmental diagnosis, Black children were more likely to receive a neurodevelopmental disorder diagnosis (adjusted odds ratio [aOR]=3.36) than children of other racial backgrounds. Moreover, no Black students received an anxiety disorder diagnosis.

Table 1.

Predictors of Diagnoses in Hospitalized Children (ages 6–12) (n=120)

Disorder IV aOR 95% CI
Depressive Sex (male) 1.086 0.438–2.695
Race (Black/African American) 0.541 0.216–1.354
Age 1.319* 1.017–1.710
Suicidal Ideation 2.438 0.953–6.242
Aggression/behavioral dysregulation 0.642 0.232–1.775
Depression/anxiety/trauma 1.741 0.644–4.705
Neurodevelopmental Sex (male) 1.982 0.678–5.789
Race (Black/African American) 3.361* 1.267–8.917
Age 0.831 0.614–1.125
Suicidal Ideation 0.962 0.351–2.641
Aggression/behavioral dysregulation 2.637 0.913–7.613
Depression/anxiety/trauma 0.324 0.090–1.172
Anxiety1 Sex (male) 0.798 0.288–3.049
Race (Black/African American) – –
Age 1.005 0.717–1.566
Suicidal Ideation 0.646 0.242–2.349
Aggression/behavioral dysregulation 0.446 0.129–2.266
Depression/anxiety/trauma 0.460 0.131–2.300
Trauma and Stress Sex (male) 1.441 0.593–3.499
Race (Black/African American) 0.866 0.333–2.251
Age 0.940 0.728–1.213
Suicidal Ideation 1.553 0.570–4.230
Aggression/behavioral dysregulation 1.265 0.454–3.524
Depression/anxiety/trauma 0.574 0.227–1.449
Open in a new tab

Notes. IV=independent variable; aOR=adjusted odds ratio.

*

p<.05

1

Because no Black students received a diagnosis with anxiety, race was not included in this analysis.

Results from the sensitivity analyses that excluded 12 year old children and focused on children between ages 6–11 also revealed race as a significant predictor of receiving a Neurodevelopmental diagnosis. Note, however, that age was no longer predictive of children receiving a depressive disorder (see Table S4 in Supplementary Materials).

Differences between Hospitalized Pre-adolescent Children and Adolescents

Differences in Race and Sex.

In order to examine if hospitalized children with a suicide-related billing code were more likely to be Black or male than adolescents, two separate logistic regressions with age group (child or adolescent) as the independent variable and sex (male or female) and race (Black or other) as dependent variables were conducted. Results were significant for race (OR=1.99, 95% Confidence Interval [CI]:1.17,3.38) and sex (OR=1.94, 95% CI:1.28,2.95), indicating that the child age group was nearly twice as likely to include boys or Black youth compared to the adolescent age group.

Comparable findings emerged from sensitivity analyses excluding 12 year old children (i.e., comparing children ages 6–11 and adolescents ages 13–18). More specifically, both models (race [n=429; OR = 2.129, 95% CI, 1.138, 3.984] and sex [n=454; OR = 2.753, 95% CI: 1.653, 4.585]) remained significant.

Differences in Diagnosis.

To test our hypothesis that children would be more likely to present with neurodevelopmental and trauma and stress related disorders, and less likely to present with depressive and anxiety disorders than adolescents, a series of logistic regressions with age group (child or adolescent) as the independent variable, controlling for sex and race, and diagnostic categories as dependent variables were run. Results, shown in Table 2, indicate that children were more likely to present with neurodevelopmental (aOR=2.99) and trauma and stress related (aOR=2.64) disorders and less likely to present with depressive disorders (aOR=0.43) than adolescents. Results by age group were not significantly different for anxiety disorders.

Table 2.

Age Group as a Predictor of Diagnoses for children ages 6–12 and adolescents ages 13–18 (n=471)

Disorder IV aOR 95% CI
Depressive Sex (male) 0.629* 0.400–0.920
Race (Black) 0.546* 0.311–0.874
Age-group (child) 0.425* 0.264–0.684
Neurodevelopmental Sex (male) 2.731* 1.578–4.729
Race (Black) 2.817* 1.510–5.255
Age-group (child) 2.985* 1.718–5.186
Anxiety Sex (male) 0.893 0.549–1.450
Race (Black) 0.267* 0.103–0.693
Age-group (child) 0.916 0.523–1.603
Trauma and Stressor Related Sex (male) 1.065 0.669–1.693
Race (Black) 0.953 0.511–1.779
  Age-group (child) 2.639* 1.405–4.262
Open in a new tab

Notes. IV=independent variable; aOR=adjusted odds ratio.

*

p<.05

Both sex and race significantly predicted depressive disorder (aOR=0.63 and aOR=0.55, respectively) and neurodevelopmental disorder (aOR=2.73 and aOR=2.82, respectively) diagnosis, with being male and Black associated with reduced odds for receiving a depressive diagnosis and increased odds for receiving a neurodevelopmental diagnosis. Black children and adolescents were significantly less likely to receive an anxiety disorder diagnosis compared to youth from other racial backgrounds (aOR=0.27).

All previously significant predictor variables remained significant in the sensitivity analyses comparing pre-adolescent children (ages 6–11) to adolescents (ages 13–18), with the exception of race in the model exploring age group as a predictor of depressive disorder diagnoses (see Table S5 in Supplementary Materials).

Discussion

To date, the only study to explore differences in children and adolescents hospitalized for STB was conducted in Ireland over a decade ago (Sarkar et al., 2010). Since then, increases in psychiatric hospitalizations and suicide have been observed among children and adolescents in the U.S. (Centers for Disease Control and Prevention, 2020; Curtin et al., 2016; Plemmons et al., 2018). Therefore, the current study aimed to contribute to a notable gap in the literature by characterizing hospitalized children with suicide-related billing codes in one U.S. psychiatric hospital and comparing demographic characteristics and diagnoses of children and adolescents.

Characteristics of Children Hospitalized for Suicide Risk

Children presented with multiple reasons for hospitalization, including suicidal ideation, reports of extreme aggression or defiance, and symptoms related to depression and anxiety. Diagnoses provided at discharge generally aligned with these symptoms, with depressive, neurodevelopmental, trauma and stressor-related, and anxiety disorders most common similar to findings reported by previous studies (Martin et al., 2016; Whalen et al., 2015; Wyman et al., 2009).

Common environmental stressors recorded in discharge reports included problems related to school, discord in the family, and interpersonal difficulties. These results are consistent with prior research, which suggests that STB, externalizing concerns, and school events (including bullying victimization), often precipitate inpatient admission for children and may serve as proximal or near-term risk factors for suicide (Dervic et al., 2008; Pompili et al., 2005; Sarkar et al., 2010). Findings point to the importance of not only linking families to care, but also intervening within school settings.

Correlates of Diagnosis within Children Hospitalized for Suicide Risk

We explored whether being male and of younger age was positively related to being diagnosed with neurodevelopmental disorders and negatively related to being diagnosed with an internalizing condition (e.g., depression, anxiety). Although bivariate correlations revealed that age (in years) and male sex was significantly associated with depressive and neurodevelopmental disorders in the expected direction, only depression remained significant in adjusted regression models controlling for key demographic variables. Boys and girls were relatively evenly represented in children ages 6–12, but the small number of young children (less than 25% were ages 6–9 years) may have limited power to detect significant effects of developmental groups.

Results also supported race as a predictor of neurodevelopmental and anxiety disorder diagnoses. Black children were significantly more likely to receive a neurodevelopmental disorder diagnosis. Moreover, no Black children in the sample received an anxiety disorder diagnosis. Because comparable findings emerged in the entire sample of both children and adolescents, this finding is discussed in greater detail in the subsequent section.

Differences between Pre-Adolescent Children and Adolescents Hospitalized for Suicide Risk

The secondary aim of the present study was to examine differences between 121 hospitalized pre-adolescent children and 381 hospitalized adolescents. With the exception of anxiety, results supported hypotheses, as hospitalized children were more likely to be Black, male, and receive a discharge diagnosis of a neurodevelopmental or trauma and stressor-related disorder and less likely to receive a depressive disorder diagnosis, compared to adolescents. These findings reinforce a pattern emerging across the literature: younger hospitalized children for suicide-related concerns appear more likely to be Black, male, and have symptoms related to neurodevelopmental disorders such as ADHD (Martin et al., 2016; Sarkar et al., 2010; Whalen et al., 2015).

Present findings are also consistent with those found in general populations demonstrating that children ages 6–11 are significantly less likely to be diagnosed with depression or anxiety, but more likely to be diagnosed with behavioral or conduct problems, than children and adolescents age 12–17 (Ghandour et al., 2019). One explanation for these diagnostic differences relates to referral systems for care, which may depend on caregivers and teachers observing indicators of risk (e.g., behavioral problems). Teachers and caregivers may be less able to identify the subtle signs of children with suicide risk struggling with internalizing symptoms. Indeed, depressive disorders in children appear to have more variable symptomology than in older children and their identification often requires outside observations as opposed to self-report (Goldman, 2012). It remains unclear if depressive disorders are less prevalent in children or if they are simply harder to identify.

Contrary to our hypothesis, significant differences between age groups were not supported for anxiety diagnoses. In one of the only studies to explore anxiety in relation to STB in clinical samples of children, significant differences between children with and without ideation were not supported for symptoms of specific phobia, separation anxiety, or generalized anxiety (Wyman et al., 2009). A recent meta-analysis suggests that the link between anxiety disorders and suicide risk in adolescent and adult populations is relatively weak (Bentley et al., 2016), perhaps explaining the lack of differences between age groups.

Findings from the present study contribute new information about the characteristics associated with hospitalization for suicide-related behavior in children: compared to adolescents, children hospitalized for suicide risk appear more likely to include higher a proportion of Black children. This finding aligns with recent research suggesting Black children are more at risk of dying by suicide (Sheftall et al., 2016), and adds to our growing knowledge of treatment-seeking in Black youth. Although trends in treatment-seeking for STB have been increasing across all ethnic and racial groups, non-Hispanic Black and Hispanic youth have demonstrated the lowest increase (Plemmons et al., 2018) and Black adolescents with suicidal ideation are significantly less likely to receive treatment compared to White families (Cummings et al., 2010). In light of the current study’s findings, studies exploring differences in treatment-seeking should stratify rates based on age in addition to ethnicity and race.

Black children and adolescents were also more likely to receive neurodevelopmental disorder diagnoses and less likely to receive internalizing disorder diagnoses (i.e., depression and anxiety). Prior research has shown that Black children and adolescents are more likely to receive behavioral, conduct, and learning disorder diagnoses and less likely to receive anxiety disorder diagnoses than White youth (Ghandour et al., 2019; Zablotsky & Alford, 2020). Because Black youth are less likely to receive psychiatric services, it is also possible that the children who are hospitalized for suicide-risk are more likely to display obvious and extreme behaviors (i.e., problem behaviors associated with neurodevelopmental disorders and more specifically, externalizing symptoms associated with ADHD) than others. The differing rates of behavioral and anxiety disorders may reflect racial bias in diagnostic assessment, with strong evidence that clinicians hold negative implicit racial bias against Black patients (Dovidio & Fiske, 2012; Van Ryn & Saha, 2011). Speculations about the reasons prevalence rates of diagnoses differ for Black individuals include service, cultural, and language barriers (Zablotsky & Alford, 2020). Thus, the differences reported here may reflect an interplay of cultural variability, differential health access, and racial bias, contributing to racial inequity at large.

Limitations

Several limitations should be considered when interpreting the findings from this study. Cases for the present study were extracted from one hospital and may be more reflective of the locale in which the study took place, limiting the external validity of findings. Because variables included in the models were based on data available in discharge summaries, they reflect clinical impressions from one time-point and causality cannot be inferred. Instead, results point to areas of focus (e.g., race and school stressors) for future longitudinal inquiries examining predictors of suicide risk in children. Future research that uses research-based measures assessing both severity and intensity of impairments, and also considers the influence of comorbidities, will provide a clearer picture of suicide risk in children.

It is also important to acknowledge that the sample size of this study was based on existing medical chart records. Thus, it is unknown if the lack of association for some of our findings within children are due to a biased sample, lack of power, or admission and diagnostic practices that differ from those used in other studies of hospitalized youth relying on medical records. Because of our sample size limitations, we elected to examine diagnoses according to broad DSM-V categories and it is possible we missed important differences within each category (e.g., differentiating between ADHD and Autism Spectrum Disorder may result in different findings). In particular, because majority of children receiving a diagnosis of neurodevelopmental disorder appeared to have ADHD, future research focused on ADHD is warranted. Moreover, only 5% of our child sample received a Disruptive, Impulsive-Control, and Conduct Disorder diagnosis, so we were not able to explore demographic differences for ODD or Conduct Disorder.

A final limitation relates to the way we defined age groups for pre-adolescent children (ages 6–12) and adolescents (ages 13–18). Although childhood and adolescence reflect distinct developmental periods, the age at which children transition to adolescence is highly variable (Euling et al., 2008). Results from sensitivity analyses conducted with pre-adolescent children ages 6–11 (excluding children age 12) revealed similar findings to main analyses that were conducted with pre-adolescent children ages 6–12; however, it is possible that children approaching puberty should be considered separately from pre-adolescent and adolescent children. Longitudinal research that includes more robust samples of children ages 6–12 should explore differences within this age group to identify changes in patterns of risk factors across development.

Conclusion

Across aims, the present findings provide a clearer picture of characteristics and features of children hospitalized for suicide risk. With the exception of depression, age and sex did not differentially predict diagnoses within pre-adolescent children; however, both age and sex did predict likelihood for diagnoses of depressive, neurodevelopmental, and trauma and stress-related disorders when compared between children and adolescents. In other words, within this sample of hospitalized children, most diagnoses were comparable across children ages 6–12, but in the entire sample of children and adolescents, diagnoses were significantly related to age group (6–12 compared to 13–18 years). An important exception to this pattern was that race differentiated diagnoses both within children and the entire sample, and results suggest a higher rate of Black children being hospitalized for suicide risk than adolescents. Considering the over-identification of behavioral disorders and under-identification of internalizing disorders in Black youth, as well as the ongoing health disparities for Black individuals in the U.S., this work must be approached with cultural sensitivity and move beyond problem identification towards strength-based frameworks for enhancing resiliency in Black children and families (Chu et al., 2010).

Supplementary Material

Supplementary Materials

Key Practitioner Message:

  • Rates of suicide and suicide-related thoughts and behaviors in pre-adolescent children have increased in recent years.

  • This study presents characteristics of pre-adolescent (ages 6–12) children hospitalized for suicide-related thoughts and behaviors and compares demographic and diagnostic differences between children and adolescents.

  • Findings reinforce the importance of proactively identifying Black boys at risk for suicide, with culturally sensitive identification and referral processes that supports linkage to care.

Acknowledgments

The project described was supported by the National Center for Advancing Translational Sciences (NCATS), National Institutes of Health, through Grant Award Number UL1TR002489. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Footnotes

Study procedures were approved by the Institutional Review Board. The authors have no conflicts of interest to report.

References

  1. Ayer L, Colpe L, Pearson J, Rooney M, & Murphy E (2020). Advancing research in child suicide: A call to action. [DOI] [PubMed] [Google Scholar]
  2. Barrio CA (2007). Assessing suicide risk in children: Guidelines for developmentally appropriate interviewing. Journal of Mental Health Counseling, 29(1), 50–66. [Google Scholar]
  3. Bentley KH, Franklin JC, Ribeiro JD, Kleiman EM, Fox KR, & Nock MK (2016). Anxiety and its disorders as risk factors for suicidal thoughts and behaviors: A meta-analytic review. Clinical Psychology Review, 43, 30–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brent DA, Baugher M, Bridge J, Chen T, & Chiappetta L (1999). Age-and sex-related risk factors for adolescent suicide. Journal of the American Academy of Child & Adolescent Psychiatry, 38(12), 1497–1505. [DOI] [PubMed] [Google Scholar]
  5. Bridge JA, Asti L, Horowitz LM, Greenhouse JB, Fontanella CA, Sheftall AH, Kelleher KJ, & Campo JV (2015). Suicide trends among elementary school–aged children in the United States from 1993 to 2012. JAMA Pediatrics, 169(7), 673–677. [DOI] [PubMed] [Google Scholar]
  6. Centers for Disease Control and Prevention. (n.d.). Web-based injury statistics query and reporting system (WISQARS). National Center for Injury Prevention and Control, Centers for Disease Control and Prevention (Producer). Http://Www.Cdc.Gov/Ncipc/Wisqars (Accessed 15 May 2020). [Google Scholar]
  7. Chu JP, Goldblum P, Floyd R, & Bongar B (2010). The cultural theory and model of suicide. Applied and Preventive Psychology, 14(1–4), 25–40. [Google Scholar]
  8. Cummings JR, Ponce NA, & Mays VM (2010). Comparing racial/ethnic differences in mental health service use among high-need subpopulations across clinical and school-based settings. Journal of Adolescent Health, 46(6), 603–606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Curtin SC, Warner M, & Hedegaard H (2016). Suicide rates for females and males by race and ethnicity: United States, 1999 and 2014. National Centre for Health Statistics, 5. [Google Scholar]
  10. Dervic K, Brent DA, & Oquendo MA (2008). Completed suicide in childhood. Psychiatric Clinics of North America, 31(2), 271–291. [DOI] [PubMed] [Google Scholar]
  11. Diagnostic A (2013). Statistical manual of mental disorders,(DSM-5) Washington. DC: Author. [Google Scholar]
  12. Dovidio JF, & Fiske ST (2012). Under the radar: How unexamined biases in decision-making processes in clinical interactions can contribute to health care disparities. American Journal of Public Health, 102(5), 945–952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Euling SY, Herman-Giddens ME, Lee PA, Selevan SG, Juul A, Sørensen TI, … & Swan SH (2008). Examination of US puberty-timing data from 1940 to 1994 for secular trends: panel findings. Pediatrics, 121(Supplement 3), S172–S191. [DOI] [PubMed] [Google Scholar]
  14. Ghandour RM, Sherman LJ, Vladutiu CJ, Ali MM, Lynch SE, Bitsko RH, & Blumberg SJ (2019). Prevalence and treatment of depression, anxiety, and conduct problems in US children. The Journal of Pediatrics, 206, 256–267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Goldman S (2012). Developmental epidemiology of depressive disorders. Child and Adolescent Psychiatric Clinics, 21(2), 217–235. [DOI] [PubMed] [Google Scholar]
  16. Corp IBM. (2019). IBM SPSS for Macintosh (Version 26) [Computer software].
  17. Lin F-G, Lin J-D, Hsieh Y-H, & Chang C-Y (2014). Quarrelsome family environment as an enhanced factor on child suicidal ideation. Research in Developmental Disabilities, 35(12), 3245–3253. [DOI] [PubMed] [Google Scholar]
  18. Luby JL, Whalen D, Tillman R, & Barch DM (2019). Clinical and psychosocial characteristics of young children with suicidal ideation, behaviors, and nonsuicidal self-injurious behaviors. Journal of the American Academy of Child & Adolescent Psychiatry, 58(1), 117–127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Martin SE, Liu RT, Mernick LR, DeMarco M, Cheek SM, Spirito A, & Boekamp JR (2016). Suicidal thoughts and behaviors in psychiatrically referred young children. Psychiatry Research, 246, 308–313. [DOI] [PubMed] [Google Scholar]
  20. Mayes SD, Fernandez-Mendoza J, Baweja R, Calhoun S, Mahr F, Aggarwal R, & Arnold M (2014). Correlates of suicide ideation and attempts in children and adolescents with eating disorders. Eating Disorders, 22(4), 352–366. [DOI] [PubMed] [Google Scholar]
  21. Miller DN (2019). Suicidal behavior in children: Issues and implications for elementary schools. Contemporary School Psychology, 23, 357–366. [Google Scholar]
  22. Mishara BL (1999). Conceptions of death and suicide in children ages 6–12 and their implications for suicide prevention. Suicide and Life-Threatening Behavior, 29(2), 105–118. [PubMed] [Google Scholar]
  23. Muthén L, & Muthén B (2017). Mplus (Version 8)[computer software].(1998–2017). Los Angeles, CA: Muthén & Muthén. [Google Scholar]
  24. Pfeffer CR, Conte HR, & Plutchik R (1980). Suicidal behavior in latency-age children: An outpatient population. Journal of the American Academy of Child Psychiatry, 19(4), 703–710. [DOI] [PubMed] [Google Scholar]
  25. Pfeffer CR, Klerman GL, Hurt SW, Kakuma T, Peskin JR, & Siefker CA (1993). Suicidal children grow up: Rates and psychosocial risk factors for suicide attempts during follow-up. Journal of the American Academy of Child & Adolescent Psychiatry, 32(1), 106–113. [DOI] [PubMed] [Google Scholar]
  26. Plemmons G, Hall M, Doupnik S, Gay J, Brown C, Browning W, Casey R, Freundlich K, Johnson DP, & Lind C (2018). Hospitalization for suicide ideation or attempt: 2008–2015. Pediatrics, 141(6), e20172426. [DOI] [PubMed] [Google Scholar]
  27. Pompili M, Mancinelli I, Girardi P, Ruberto A, & Tatarelli R (2005). Childhood suicide: A major issue in pediatric health care. Issues in Comprehensive Pediatric Nursing, 28(1), 63–68. [DOI] [PubMed] [Google Scholar]
  28. Ridge Anderson A, Keyes GM, & Jobes DA (2016). Understanding and treating suicidal risk in young children. Practice Innovations, 1(1), 3. [Google Scholar]
  29. Sarkar M, Byrne P, Power L, Fitzpatrick C, Anglim M, Boylan C, & Morgan S (2010). Are Suicidal Phenomena in Children Different to Suicidal Phenomena in Adolescents? A Six-Year Review. Child and Adolescent Mental Health, 15(4), 197–203. [DOI] [PubMed] [Google Scholar]
  30. Scheeringa MS (2016). Validity of measurement of suicidal ideas in very young children. Journal of the American Academy of Child & Adolescent Psychiatry, 55(3), 243. [DOI] [PubMed] [Google Scholar]
  31. Sheftall AH, Asti L, Horowitz LM, Felts A, Fontanella CA, Campo JV, & Bridge JA (2016). Suicide in Elementary School-Aged Children and Early Adolescents. Pediatrics, 138(4), 1–10. aph. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Tishler CL, Reiss NS, & Rhodes AR (2007). Suicidal behavior in children younger than twelve: A diagnostic challenge for emergency department personnel. Academic Emergency Medicine, 14(9), 810–818. [DOI] [PubMed] [Google Scholar]
  33. Tossone K, Jefferis E, Bhatta MP, Bilge-Johnson S, & Seifert P (2014). Risk factors for rehospitalization and inpatient care among pediatric psychiatric intake response center patients. Child and Adolescent Psychiatry and Mental Health, 8(1), 27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Van Ryn M, & Saha S (2011). Exploring unconscious bias in disparities research and medical education. JAMA, 306(9), 995–996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Whalen DJ, Dixon-Gordon K, Belden AC, Barch D, & Luby JL (2015). Correlates and consequences of suicidal cognitions and behaviors in children ages 3 to 7 years. Journal of the American Academy of Child & Adolescent Psychiatry, 54(11), 926–937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wyman PA, Gaudieri PA, Schmeelk-Cone K, Cross W, Brown CH, Sworts L, West J, Burke KC, & Nathan J (2009). Emotional triggers and psychopathology associated with suicidal ideation in urban children with elevated aggressive-disruptive behavior. Journal of Abnormal Child Psychology, 37(7), 917–928. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Zablotsky B, & Alford JM (2020). Racial and Ethnic Differences in the Prevalence of Attention-Deficit/Hyperactivity Disorder and Learning Disabilities among US Children Aged 3–17 Years. NCHS Data Brief. No. 358. National Center for Health Statistics. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Materials
NIHMS1718193-supplement-Supplementary_Materials.docx (34.5KB, docx)

RESOURCES