Characteristics of Fatalities Among Sexual- and Gender-Diverse Youth in the United States: 2015–2020

This study used multistate data from CDR conducted in the United States. Although CDR programs vary, CDR teams often include representation from: medical examiner/coroner, law enforcement, first responders, health care providers, social services, public health, prosecuting attorney, and others.^22,23 CDR teams gather detailed information about the fatality from multiple sources, identifying relevant risk and protective factors, and informing future prevention recommendations.^22–24 CDR protocols regarding which deaths are reviewed vary by state.²³ 

In 2005, the NFR-CRS was launched as a web-based standardized data entry tool for local and state CDR programs in the United States.²⁵ The National Center for Fatality Review and Prevention (National Center) maintains, oversees, and disseminates data from the NFR-CRS and provides support and trainings to CDR teams. More than 2600 data elements describing the child and the child’s death can be reported in the database, including demographic, injury, circumstance, investigation, and prevention recommendation variables.^26,27 As of June 2023, 47 states (excluding Maine, North Carolina, and Vermont) and the District of Columbia use the NFR-CRS for CDR reporting.²⁸ 

Versions 1.0 through 4.1 of the NFR-CRS (approximately January 2005 through April 2018) included 2 questions related to gender identity and sexual orientation: (1) “If child was over age 12, what was the child’s gender identity?” with response options “male, female, unknown,” and (2) “If child over age 12, what was the child’s sexual orientation?” with response options “heterosexual, gay, lesbian, bisexual, questioning, unknown.” In NFR-CRS, version 5.0, beginning April 2018, these questions were removed from the NFR-CRS. Then, in version 5.1 (April 2020), the NFR-CRS added these questions back in, with revised response options (gender identity options: “no identity expressed,” “male – not transgender,” “female – not transgender,” “transgender male,” “transgender female,” “non-binary,” “other, specify,” “unknown”; sexual orientation options: “no orientation expressed,” “straight/heterosexual,” “gay/lesbian,” “bisexual,” “questioning,” “other, specify,” “unknown”), which are asked of decedents older than 1 year of age. No data from earlier versions of these questions were cross walked at this time, however, multiple states updated older deaths. These questions have been asked in their current form since version 5.1 (NFR-CRS is currently in version 6.0) and are what we used for SGD selection criteria for this study.²⁹ 

We requested data on fatalities among youth ages 0 to 19 years who died between January 1, 2015, and December 31, 2020. Because the NFR-CRS does not have a specific variable to identify all SGD youth, fatalities meeting our study criteria were identified using the variables “child’s gender identity” and “child’s sexual orientation.” For this study, a child was considered SGD if their gender identity was reported as “transgender male,” “transgender female,” “nonbinary,” “other, specify,” or if the child’s sexual orientation was reported as “gay/lesbian,” “bisexual,” “questioning,” or “other, specify.” The “other, specify” variable allows for free text responses and was selected to capture gender identities and sexual orientations that do not fall within the response options provided. See Supplemental Table 4 for gender identity and sexual orientation frequencies among the SGD youth.

We also requested data for fatalities among youth who were not identified as SGD for comparison. The comparison group was defined as having a reported gender identity as “male, not transgender” or “female, not transgender” and a reported sexual orientation of “straight/heterosexual.” To identify matches, race categories were collapsed into 3 groups: Black, White, Other. Each SGD death was randomly matched to 2 non-SGD deaths by age (in years) and race. One age/race combination did not have enough non-SGD matches and was modified as follows. There were only 8 available exact non-SGD matches for 5, 15-year-old/Other SGD deaths. To bridge this gap in matching, 2 16-year-old non-SGD/Other deaths were included in this combination. Identification of SGD deaths and matched deaths was completed by the Data Team at the National Center due to restrictions on releasing narrative data. The full “other, specify” responses were reviewed by the Data Team at the National Center when identifying SGD deaths. This study team received the first 30 characters of these responses.

To comprehensively describe fatalities among SGD youth, we requested child demographic variables, incident and investigation variables, child history, and life stressor variables. Characteristics of suicide-related fatalities among SGD and non-SGD youth were compared using NFR-CRS variables specific to this manner of death. Race categories were collapsed into Black, White, and Other as described previously to facilitate matching. Race categories in “Other” include American Indian, Alaska Native, Asian, Native Hawaiian, Pacific Islander, and Multiracial. Race and ethnicity, which represent social constructs, are collected in the NFR-CRS as reported on the death certificate. Differences among stratified gender identities and sexual orientations of SGD youth are described using several important variables. However, because of small numbers, the SGD deaths were aggregated into 1 category for further analyses and comparisons to the matched group.

Data analysis was conducted using SPSS 28.0 for Macintosh (IBM Corporation, Armonk, NY). Basic descriptive statistics are reported with frequencies and percentages. χ² tests were performed to compare proportions of characteristics between SGD and non-SGD fatalities, with P values ≤ .05 indicating statistical significance. Responses to variables that were “unknown” or “not specified” were combined into “Missing/Unknown” and excluded from the denominator when calculating percentages. Some variables included a “Not Applicable” response option. These frequencies are reported but excluded from the denominator when calculating percentages. Some variables (eg, life stressors) included response options of “yes” or “not specified.” For these variables, only the “yes” responses are reported. Following National Center data use protocols, all cell counts ≤5 were suppressed to prevent inadvertent identification of decedents. Variables with ≥50% missing data were not reported. This study was exempt from review by the authors’ institutional review board. This study was approved by the National Center, but the content does not necessarily represent the official views of the National Center, its funders, or participating states.

We received 2015–2020 death data for 179 fatalities among SGD youth and 358 fatalities among non-SGD matches from the National Center. The study team excluded 3 deaths because SGD status could not be determined despite meeting selection criteria (eg, individual’s sex was male, “unknown” gender identity, and sexual orientation as “other, specify” with the response stating they had sex with females). In total, this study included 176 SGD deaths and 352 non-SGD matched comparison deaths.

Among SGD deaths with a known gender identity, 26.8% were identified as transgender (Supplemental Table 4). Of SGD deaths with identified sexual orientations, 45.1% were gay/lesbian and 23.2% were bisexual. SGD deaths were stratified into mutually exclusive categories of transgender (n = 30) compared with gay/lesbian/bisexual (n = 89) to examine differences. Greater proportions of transgender youth had ever been diagnosed with depression (56.7% versus 31.5%, P = .01) and/or anxiety (33.3% versus 11.2%, P = .005), compared with gay/lesbian/bisexual youth.

Among the 176 SGD youth fatalities identified during the study period, the majority of decedents were White (76.7%) and non-Hispanic (84.0%; Table 1). A greater percentage of SGD youth had received or were currently receiving mental health services, compared with non-SGD youth (68.8% versus 50.2% and 49.0% versus 31.0%, respectively; P < .001). Nearly half of SGD youth were prescribed medication for mental illness at the time of death (45.1%), compared with 27.8% of non-SGD youth (P < .001). In addition, a greater percentage of SGD youth visited an emergency department for mental health care within the 12 months before death (22.7% versus 9.7%; P = .001).

A significantly greater proportion of SGD youth had ever experienced homelessness, compared with non-SGD youth (9.8% versus 4.1%; P = .04). A greater proportion of SGD youth were identified as having experienced problems in school, compared with non-SGD youth (62.5% versus 47.0%; P = .004). SGD youth had lower percentages of reported criminal/delinquent history and substance use/abuse, compared with non-SGD youth (11.6% versus 22.2% and 31.9% versus 45.4%, respectively; P = .008).

The most frequently documented manner of death among SGD youth was suicide (81.4%; Table 2). There were 140 suicide deaths among SGD youth and 183 suicide deaths among non-SGD youth (Table 3). Age and race distributions were not statistically different between the 2 groups because comparison group was matched by age and race. Among decedents who had experienced the death of a peer, friend, or family member, for SGD youth, a greater percentage of those fatalities resulted from suicide (66.7% versus 34.6%; P = .03). More SGD youth were identified as having ever had a diagnosis of a depressive disorder, compared with non-SGD youth (44.3% versus 29.0%; P = .02). Reported past suicidal behavior/attempt history was not statistically different between SGD and non-SGD youth. Gender identity and sexual orientation were reported as stressors for 24.3% and 32.9% of SGD youth. SGD youth were more frequently reported as having been a victim of bullying, compared with non-SGD youth (27.9% versus 14.8%; P = .004).

This study characterized fatalities among SGD youth and compared them to fatalities among non-SGD youth using 6 years of multistate data from the NFR-CRS. The SGD status of children and adolescents is often not reported in morbidity and mortality data, despite experiencing discrimination and stigma that may lead to stressful social environments and negatively impact health.^5,30 This study provides insight into the risk factors associated with premature death, particularly suicide, among this understudied population.

A greater proportion of deaths were attributed to suicide among SGD youth, compared with non-SGD youth. Several factors were identified that may have contributed to these fatalities. Among SGD youth who died by suicide, 44.3% were reported to have been diagnosed with a depressive disorder. This is consistent with prior research indicating 55.1% of surveyed SGD youth screened positive for depression.¹ In the present study, a greater proportion of SGD youth suicide decedents were reported to have been a victim of bullying compared with non-SGD youth who died by suicide. It is well documented that SGD youth experience more bullying and harassment than their non-SGD peers, and most do not report this victimization to an adult.^1,31 Most antibullying interventions are implemented through schools, reflecting a critical place of needed support for SGD youth.³² There are a number of strategies schools should adopt to create supportive climates for SGD youth, including: supportive peer network groups; antidiscrimination polices explicitly protecting SGD identities; policies supporting use of chosen name and pronouns; trained and supportive school staff; and SGD-inclusive curricula. However, successful implementation of these strategies can be difficult, particularly in communities where the political climate is hostile to SGD people.^33–35 Rigorous research is needed to identify effective interventions to prevent bullying and address its impact on SGD youth.^5,36 SGD and non-SGD suicide decedents were reported to have experienced life stressors such as family discord, parental divorce/separation, and arguments with parents with similar frequency. Although parental rejection of SGD youth is a widely reported and harmful practice, it could not be examined as a risk factor for suicide in the present study.^1,37–39 

Nearly 70% of SGD decedents had received prior mental health services, significantly more than among non-SGD decedents. SGD youth also had significantly higher percentages of current mental health services, mental health–related emergency department visits, and current prescribed medication for mental illness, compared with non-SGD youth. Understanding the types of mental health services SGD youth are accessing and their effectiveness is imperative. For example, gender-affirming care is associated with reduced risk of suicidality, self-harm, and emotional and behavior problems for transgender youth, but access is threatened by state-level efforts to ban this care.^8,40,41 The American Academy of Pediatrics affirms the importance of gender-affirming care and pediatric healthcare environments that are welcoming of SGD youth.^42,43 More empirical evidence on the efficacy of and access to mental health treatment of SGD youth is needed, and these findings support prioritizing those efforts.^44,45 A holistic approach is also important, including interventions to prevent stigmatization, improve social supports, and promote social safety in schools and society.^46,47 

There are limitations to consider for this study. Participation in the NFR-CRS is voluntary and not all US states participate. CDR team protocols vary by jurisdiction and resources. Not all child deaths are reported to NFR-CRS, and the quality of data input by CDR teams may vary, although the National Center provides training and technical assistance to states and monitors data quality.²⁴ These data are not representative of the general US population and cannot be used to calculate rates or trends over time. It is not advisable to make inferences from variables with high proportions of missing/unknown data; thus, some variables could not be analyzed (>50% missing data). In the interest of sharing as much as possible about this understudied population, some analyses were conducted using variables with high proportions of missing data. However, they may not be representative of the whole group and inferences should not be drawn. Although many SGD youth were reported to have received mental health services, details about the timing, frequency, and quality of care could not be determined. Some variables, such as some life stressors, were not collected for all manners of death or were introduced in later versions of the CDR form. These limitations had to be considered when deciding which life stressor variables could be examined and may have resulted in some life stressor responses being underreported.

Additionally, research has shown that youth most often disclose their SGD identities to their friends and classmates before disclosing their SGD status with individuals who would contribute information about the child to CDR (eg, parents, school staff, medical providers), resulting in a risk of children’s identities being misclassified.¹ In the absence of self-identification and consensus on the most accurate way to capture gender identity and sexual orientation of decedents during death investigations, variations among CDR programs and differing levels of acceptance of SGD youth throughout the United States introduce bias and exacerbate underrepresentation of this group. Although we combined sexual and gender diverse youth for most analyses because of the small sample size, sexual orientation and gender identities should be stratified, when possible, to better understand unique stressors and inform care.

Despite these limitations, NFR-CRS offers a unique opportunity to examine characteristics of deaths among SGD youth that cannot be studied with other datasets. Additional improvements should be made to this and other public health surveillance systems to improve researchers’ ability to quantify morbidity and mortality among SGD individuals, as well as relevant risk and protective factors that can inform future prevention efforts.

SGD people are disproportionately burdened by risk factors for preventable death. In this study, most SGD fatalities occurred by suicide, shedding light on the grave threat this presents to SGD youth. Although many SGD youth received at least some mental health treatment, tailored interventions and monitoring by informed and affirming health care providers are needed, as well as a wider focus on the social and environmental stressors SGD youth face. Pediatric health care providers must not only pay special attention to the environment of their SGD patients, but also advance cultural acceptance and access to care by being strong advocates of supportive policy initiatives and speaking out against harmful legislation. When referring to mental health services, referring providers must be mindful of the accepting providers’ commitment to offering welcoming and gender-affirming care. Further research and advocacy are necessary to develop and promote interventions which foster social safety and acceptance of SGD youth in homes, schools, and society.

The authors are grateful for the work of Esther Shaw, MSIS, and the staff at the National Center for their support and assistance for this study. This work was completed while Bridget Duffy was a Master of Public Health student at the Johns Hopkins Bloomberg School of Public Health and a Fellow of the Bloomberg American Health Initiative.

CDR

child death review

National Center

National Center for Fatality Review and Prevention

NFR-CRS

National Fatality Review-Case Reporting System

SGD

sexual and gender diverse