Lesbian, gay, bisexual, transgender, and queer (LGBTQ) youth are more likely to be bullied and to report suicidal thoughts and behaviors than non–LGBTQ youth.1 Whether bullying is a more common antecedent among LGBTQ youth who die by suicide, however, is unknown. We investigated this question using postmortem records from the 2003-2017 National Violent Death Reporting System (NVDRS).
Details on the NVDRS can be found on the Centers for Disease Control and Prevention website.2 The NVDRS is compiled by trained abstractors in contributing state departments of health who collate information from death certificates, coroner or medical examiner records, and law enforcement reports. Each NVDRS record includes 2 narratives summarizing the coroner or medical examiner records and law enforcement reports describing suicide antecedents as reported by the decedent’s family or friends; the decedent’s diary, social media, and text or email messages; and any suicide note. We classified these narratives for LGBTQ status and bullying among 9884 decedents aged 10 to 19 whose narratives included any coroner or medical examiner records or law enforcement report information from January 1, 2003, through December 31, 2017. Given the use of data solely from decedents, this cohort study of postmortem records from the NVDRS was designated as not human subjects research by the Yale Human Subjects Committee.
Classifying Records for LGBTQ Status and Bullying
Codes denoting sexual orientation and transgender status were added to the NVDRS in 2013, and same-sex partnership was added in 2015. We categorized decedents precoded as lesbian, gay, bisexual, transgender, or same-sex partnered as LGBTQ (n = 194). To capture LGBTQ status for suicides that occurred before these codes (2003-2012), we adapted a process previously used by the Centers for Disease Control and Prevention to classify LGBTQ records through a full-text search of coroner or medical examiner and law enforcement narratives.3 We first developed a list of potential search terms (eg, “his boyfriend”). Using this list, we identified 176 potential LGBTQ records. We then reviewed 50 randomly selected narratives from this group to obtain additional terms and searched again, identifying a total of 251 potential LGBTQ records. We classified 140 as LGBTQ if the decedent (1) self-identified as LGBTQ; (2) was perceived to be LGBTQ by family, friends, or peers; or (3) accessed treatment for gender dysphoria (eg, hormone therapy). Reviewer pairs double-coded 50 narratives (98% agreement; κ = 0.96) and then independently coded remaining records. Overall, of the 9884 decedents, 334 (3.4%) were LGBTQ-classified.
We followed a similar procedure to classify bullying from coroner or medical examiner records and law enforcement narratives. We first developed a list of search terms for bullying, reviewed 50 randomly selected narratives identified by these terms to obtain additional terms, and searched again. The finalized list identified 601 potential records. We classified 490 of 9884 deaths (5.0%) as associated with bullying if (1) the decedent self-reported being bullied (eg, suicide note); (2) witnesses reported the decedent was bullied; or (3) postmortem investigation revealed bullying (eg, text messages). Reviewer pairs again obtained high agreement (94% agreement; κ = 0.82).
Proportions and χ2 tests documented differences between LGBTQ-classified and non–LGBTQ-classified decedents (Table). Logistic regression analyses examined the association between assigned LGBTQ status and bullying, adjusting for age, race/ethnicity, sex, depressed mood, mental health problems, family relationship problems, intimate partner problems, state of residence, and incident year. Results were evaluated at P < .05 (2-tailed). Data were analyzed from October 11, 2019, to December 12, 2019.
Table. Demographic and Antecedent Circumstance Differences Between LGBTQ-Classified and Non–LGBTQ-Classified Youth Who Died by Suicide From the National Violent Death Reporting System, 2003-2017 (N = 9884)
Of the 334 LGBTQ-classified decedents aged 10 to 19 years (mean [SD] age, 16.45 [2.02] years; legal sex, 177 [53.0%] male and 157 [47.0%] female), 69 (20.7%) were classified as being bullied, compared with 421 of 9550 (4.4%) non–LGBTQ-classified decedents (mean [SD] age, 16.70 [2.05] years; legal sex, 7300 [76.4%] male and 2250 [23.6%] female) (P < .001) (Figure). LGBTQ-classified decedents evidenced 4.92 times the odds of being bullied compared with non–LGBTQ-classified decedents (95% CI, 3.58-6.82; P < .001). Younger LGBTQ-classified decedents were at greatest risk; 21 of 31 (67.7%) LGBTQ-classified youth aged 10 to 13 years were classified as being bullied. Despite this disparity, the great majority of bullying-associated deaths were among non–LGBTQ-classified youth (85.9%, derived from 421 non–LGBTQ-classified youth with bullying-associated deaths divided by 490 bullying-associated deaths).
Figure. Prevalence of Bullying as an Antecedent to Suicide by LGBTQ Classification and Age at Death From the National Violent Death Reporting System, 2003-2017
Prevalence of bullying among decedents in the total sample was 5.0% for the total group, 20.7% for the LGBTQ-classified group, and 4.4% for the non-LGBTQ-classified group. Prevalence of bullying among decedents aged 10 to 13 years was 17.0% for the total group, 67.6% for the LGBTQ-classified group, and 15.0% in the non-LGBTQ-classified group. Prevalence of bullying among decedents aged 14 to 16 years was 7.5% for the total group, 27.8% for the LGBTQ-classified group, and 6.7% for the non-LGBTQ-classified group. Prevalence of bullying among decedents aged 17-19 years was 1.9% for the total group, 7.3% for the LGBTQ-classified group, and 1.7% for the non-LGBTQ-classified group. LGBTQ indicates lesbian, gay, bisexual, transgender, queer.
Limitations of the NVDRS detailed elsewhere include that LGBTQ information is not systematically reported in death records and is, therefore, probably underestimated.4,5 Nevertheless, our findings underscore that bullying can be a deadly antecedent to suicide, especially among LGBTQ youth. In addition to school-based interventions, pediatricians can help to reduce this risk through adopting clinical practice approaches sensitive to the vulnerabilities of LGBTQ youth.6
Accepted for Publication: January 22, 2020.
Corresponding Author: Kirsty A. Clark, PhD, MPH, Department of Social and Behavioral Sciences, Yale School of Public Health, 60 College St, New Haven, CT 06510 (email@example.com).
Published Online: May 26, 2020. doi:10.1001/jamapediatrics.2020.0940
Author Contributions: Dr Clark had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Clark, Cochran, Maiolatesi.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Clark, Cochran.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Clark, Maiolatesi.
Obtained funding: Clark.
Supervision: Cochran, Pachankis.
Conflict of Interest Disclosures: Dr Pachankis reported receiving personal fees from Oxford University Press outside the submitted work. No other disclosures were reported.
Funding/Support: This study was supported by the National Violent Death Reporting System New Investigator Award grant NU38OT000294 from the American Public Health Association funded by the Centers for Disease Control (Dr Clark). This study was partially supported by grant T32MH020031-20 from the Yale University Center for Interdisciplinary Research on AIDS training program funded by the National Institute of Mental Health (Mr Maiolatesi) and grant R21MH115344 from the National Institute for Mental Health (Dr Cochran).
Role of the Funder/Sponsor: The funders had no role in the conduct of the study; management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The Centers for Disease Control collected and provided the data and approved the study design in initial scientific review.
Disclaimer: Findings reported in this article do not reflect the views of the Centers for Disease Control and Prevention or the National Institute of Mental Health.