Abstract
Introduction:
Transgender and gender diverse people often face discrimination and may experience disproportionate emotional distress, leading to suicide attempts. It is, therefore, essential to estimate the frequency and potential determinants of suicide attempts among transgender and gender diverse individuals.
Methods:
Longitudinal data on 6,327 transgender and gender diverse individuals enrolled in three integrated healthcare systems were analyzed to assess suicide attempt rates. Incidence was compared in transmasculine and transfeminine people, by age and race/ethnicity and according to mental health status at baseline. Cox proportional hazards models examined rates and predictors of suicide attempts during follow-up. Data were collected in 2016 and analyses were conducted in 2019.
Results:
During follow-up, 4.8% of transmasculine and 3.0% of transfeminine patients had at least one suicide attempt. Suicide attempt rates were more than seven times higher among patients aged <18 years compared with those aged >45 years, more than three times higher among those with previous history of suicide ideation or suicide attempts compared with patients with no such history, and two to five times higher among those with one to two mental health diagnoses and more than two mental health diagnoses at baseline versus those with none.
Conclusions:
Among transgender and gender diverse inviduals, younger people, those with prior suicidal ideation or attempts, and those with multiple mental health diagnoses are at higher risk for suicide attempts. Future research should examine the impact of gender-affirming healthcare use on risk of suicide attempts and identify targets for suicide prevention interventions among transgender and gender diverse people in clinical settings.
INTRODUCTION
The term “transgender and gender diverse” (TGD) describes individuals who have gender identities, gender expressions, or behavior not traditionally associated with their sex assigned at birth.1 Although TGD individuals self-identify as men or women, a substantial proportion reject binary gender categories.2 To reflect the wide spectrum of gender identities, the terms “transfeminine” (TF) and “transmasculine” (TM) refer to an individual whose gender identity differs in any way from their male or female sex assigned at birth, respectively. Between 0.5% and 1% of adults in the U.S. identify as TGD,3,4 and the proportion of TGD individuals in the U.S. has been increasing.3,5 TGD people represent a sizable and growing population whose physical and mental health require more attention.
People who identify as TGD are more likely to experience depression and other mental health concerns compared with others.6–8 In an online survey of the TGD population in the U.S., 44% of respondents were found to be clinically depressed.6 Similarly, in a cohort study of more than 500 TGD individuals, nearly two thirds of TM and more than half of TF individuals were depressed.9 The lifetime prevalence of suicide attempts is estimated to be between 32% and 41% for both TM and TF populations,10–13 which is much higher than the average prevalence in the U.S. population of 1.9% to 8.7%.14
Furthermore, previous research has found that TGD people who are younger, unemployed, have lower income, or have experienced sexual coercion or gender-based violence are at higher risk for suicide attempts.11,15,16 Though much of the previous literature on suicide attempts among TGD people is limited to qualitative studies, cross-sectional analyses, or studies that are based on small convenience samples,17–19 a prior study of Veterans Health Administration records indicated that prevalence of suicide-related behaviors was up to 20 times higher among transgender veterans compared with cisgender veterans.10 In addition, social stressors (housing instability, financial strain, and experiences of violence) were associated with higher prevalence of suicidal ideation or history of suicide attempts among transgender veterans.20 Yet, studies utilizing health records for assessing incidence, rather than prevalence, of suicide attempts in the transgender population have not been conducted.
To address these knowledge gaps, suicide attempts are examined in a large longitudinal cohort of TGD people (N=6,287) enrolled in three large integrated health systems in the U.S., the largest longitudinal study of clinical experiences of TGD people conducted to date.21 The purpose of this study is to evaluate the overall incidence of suicide attempts among TGD people and compare rates of these events across demographic groups and by mental health diagnoses at baseline.
METHODS
Study Sample
The data for the current analyses originated from the Study of Transition Outcomes and Gender. This cohort was assembled from electronic medical records (EMRs) of individuals enrolled in three Kaiser Permanente Health Plan regions: Northern California, Southern California, and Georgia. The details of cohort ascertainment methodology are provided elsewhere.21 Briefly, TGD individuals were identified from the three participating sites between 2006 and 2014 based on relevant ICD-9 codes or the presence of specific keywords within the free-text sections of medical records, with follow-up extending through 2016. Each cohort member was assigned an index date—the first instance the EMR contained evidence of TGD status. Another free-text search was conducted to ascertain each person’s TF or TM status. Once the cohort was validated,21 patient identification numbers were linked to multiple data sources including ICD-9 and ICD-10 diagnostic codes and healthcare utilization records. IRBs at the three Kaiser Permanente sites and Emory University, which served as the coordinating center, approved the study.
Measures
The main outcome variable in the present analysis was EMR-based evidence of attempted suicide upon presentation to an emergency department. The event of interest was defined by ICD-9 or ICD-10 codes, which included self-inflicted injury, possible self-inflicted injury, and self-inflicted injury/poisoning (Table 1), an approach to capturing suicide attempts that has been validated previously.22,23
Table 1.
ICD 9th and 10th Edition Codes for Self-Inflicted Injury and Possible Self-Inflicted Injury
| Description | ICD-9 codes | ICD-10 codesb |
|---|---|---|
| Self-inflicted injuries | ||
| Self-inflicted poisoning by solid or liquid substances | E950.a | T36.aX2A, T40.a X2A |
| Self-inflicted poisoning by gases and vapors | E951.a, E952.a | T58.a X2A, T59.aX2A |
| Self-inflicted injury by hanging, strangulation, or suffocation | E953.a | T71.1a 2A |
| Self-inflicted injury by submersion or downing | E954.a | X71.a XXA |
| Self-inflicted injury by guns or explosives | E955.a | X72.XXXA, X73.a XXA |
| Self-inflicted injury by cutting or piercing | E956.a | X78.a XXA |
| Self-inflicted injury by jumping | E957.a | X80.XXXA |
| Self-inflicted injury by other and unspecified means | E958.a | X82.a XXA, X83.8XXA |
| Possibly self-inflicted injuries | ||
| Possibly self-inflicted injury by solid or liquid poisoning | E980.a | T36.a X4A, T40.a X4A |
| Possibly self-inflicted injury by gases and vapors | E981.a, E982.a | T58.a X4A, T59.a X4A |
| Possibly self-inflicted injury by hanging, strangulation, or suffocation | E983.a | T71.1a 4A |
| Possibly self-inflicted injury by submersion or drowning | E984.a | Y21.9XXA |
| Possibly self-inflicted injury by guns or explosives | E985.a | Y22.XXXA, Y23.a XXA |
| Possibly self-inflicted injury by cutting or piercing | E986.a | Y28.a XXA |
| Possibly self-inflicted injury by jumping | E987.a | Y30.XXA |
| Possibly self-inflicted injury by other and unspecified means | E988.a | Y33.XXXA |
Represents any digit.
Example codes for ICD-10 provided, complete list can be accessed from the Mental Health Research Network GitHub site (https://github.com/MHResearchNetwork/Diagnosis-Codes).
The demographic variables of interest included age, TM/TF status, and race/ethnicity. With respect to age, patients were categorized as 3–17, 18–25, 26–35, 36–45, and >45 years. Because there were only two events of interest among individuals aged >65 years, further subcategorization of the oldest age group was not possible. Race/ethnicity was categorized as Hispanic, non-Hispanic black, non-Hispanic white, and other/unknown. A count of mental health diagnoses by category at baseline (i.e., on or before the index date) was calculated for each individual. The categories of mental health diagnoses included: anxiety disorders, attention-deficit/hyperactivity disorders, autism spectrum disorders, bipolar disorders, depressive disorders, schizophrenia spectrum disorders, substance use/abuse, and other disorders, which included conduct/disruptive disorders, eating disorders, dementia, other psychoses, and personality disorders. For each individual, the total number of mental health diagnoses by category was categorized as zero, one to two, or more than two. Additionally, suicidal ideation (V62.84) and suicide attempts (E950–E958, E980–E988) prior to index date were ascertained and expressed as a binary (ever/never) variable. Mental health diagnoses at baseline were determined based on on ICD-9 codes only because cohort ascertainment was completed prior to 2015; that is, before ICD-10 codes were introduced at the participating sites in 2015. New events during follow-up were ascertained using both ICD-9 and ICD-10 codes through the end of 2016 (Table 1). ICD-9 codes for late effects of self-inflicted injury (E959 and E989) and ICD-10 codes reflecting a follow-up encounter for the initial event (marked by letter D as in T36.0X2D) were excluded.
The main parameter of interest was the rate of suicide attempts. Time under observation was based on health plan enrollment records; it started on the index date and ended at the time of disenrollment, death, or end of the study (December 31, 2016), whichever occurred first. Gaps in enrollment <90 days were likely due to delay in insurance renewal and do not actually result in interruption of healthcare services. For this reason, only gaps in coverage >90 days were considered as evidence of interrupted coverage, similar to previous studies.24 All patients who had at least 1 day of enrollment were included.
Statistical Analysis
The analysis involved two approaches. A Cox proportional hazards model was used to examine time to first event, where the event of interest was the first suicide attempt after the index date. Baseline covariates were measured at index date and follow-up was extended until first diagnosis of self-harm, disenrollment, death, or the end of study. Proportional hazard assumptions were violated for race/ethnicity and study site; therefore, all models were stratified on these two variables.
An extended Cox model that incorporated the counting process approach was used to account for recurrent suicide attempts.25 As the correlation of events for each cohort member is captured by specification of time-varying covariates, each study participant can contribute multiple follow-up periods with variable entry and with each follow-up extending until the next event, disenrollment, or the end of the study. The start of follow-up following each event was delayed by 1 week to exclude repeated encounters related to the same suicide attempt. Age and baseline history of suicide attempt or suicidal ideation were measured at the start of each follow-up interval. All other variables were assessed at index date. Validity of the proportional hazard assumption was examined for each model by inspecting log–log curves. The results of all time-to-event analyses were expressed as hazard ratios (HRs) and the corresponding 95% CIs. All analyses were performed using SAS, version 9.4. Data analyses were completed in 2019.
RESULTS
Selected characteristics of the Study of Transition Outcomes and Gender cohort are shown in Table 2. A total of 6,327 TGD individuals were included in the analysis. Of those, 2,875 (45%) were TM and 3,452 (55%) were TF (Table 2). The average follow-up was 4.5 years (1,627 days) among TF and 4.1 years (1,502 days) among TM study participants. A total of 248 cohort members (117 TF and 131 TM) experienced at least one suicide attempt. The total number of attempts was 166 in the TF group and 208 in the TM group. The crude rate of suicide attempts in the TF and TM study population was 3.0 (95% CI=2.5, 3.4) and 4.8 (95% CI=4.2, 5.5) per 100,000 person-days of follow up, respectively.
Table 2.
Characteristics of TM and TF Study Participants With and Without at Least One Episode of Self-Harm During Follow Up (2006–2016) (N=6,327)
| Participant characteristics | TF (N=3,452) | TM (N=2,875) | ||
|---|---|---|---|---|
| At least 1 self-harm eventa | No self-harm events | At least 1 self-harm eventa | No self-harm events | |
| n (%)b | n (%) | n (%)b | n (%) | |
| Age at index date, years | ||||
| 3–17 | 45 (38.5) | 542 (16.3) | 79 (60.3) | 663 (24.2) |
| 18–25 | 25 (21.4) | 622 (18.7) | 25 (19.1) | 718 (26.2) |
| 26–35 | 15 (12.8) | 577 (17.3) | 15 (11.5) | 689 (25.1) |
| 36–45 | 21 (18.0) | 551 (16.5) | 7 (5.3) | 339 (12.4) |
| >45 | 11 (9.4) | 1,043 (31.3) | 5 (3.8) | 335 (12.2) |
| Race/Ethnicity | ||||
| Hispanic | 22 (18.8) | 682 (20.5) | 31 (23.7) | 466 (17.0) |
| Non-Hispanic black | 15 (12.8) | 226 (6.8) | 12 (9.2) | 242 (8.8) |
| Non-Hispanic white | 67 (57.3) | 1,755 (52.6) | 74 (56.5) | 1,589 (57.9) |
| Other/Unknown | 13 (11.1) | 672 (20.2) | 14 (10.7) | 447 (16.3) |
| Study site | ||||
| KPNC | 72 (61.5) | 1,873 (56.2) | 81 (61.8) | 1,740 (63.4) |
| KPGA | 2 (1.7) | 93 (2.8) | 0 (0) | 79 (2.9) |
| KPSC | 43 (36.8) | 1,369 (41.1) | 50 (38.2) | 925 (33.7) |
| Count of mental health diagnoses at baselinec | ||||
| 0 | 28 (23.9) | 1,761 (52.8) | 18 (13.7) | 1,262 (46.0) |
| 1–2 | 50 (42.7) | 1,153 (34.6) | 60 (45.8) | 1,077 (39.3) |
| >2 | 39 (33.3) | 421 (12.6) | 53 (40.5) | 405 (14.8) |
| History of suicidal ideation or self-harm at baselinec | ||||
| Yes | 24 (20.5) | 96 (2.9) | 40 (30.5) | 154 (5.6) |
| No | 93 (79.5) | 3,239 (97.1) | 91 (69.5) | 2,590 (94.4) |
| Total | 117 (3.4) | 3,335 (96.6) | 131 (4.6) | 2,744 (95.4) |
Includes any diagnosis of self-inflicted injury, possibly self-inflicted injury, and self-inflicted injury/poisoning.
Column percentages within categories, row percentages for totals.
Baseline refers to diagnoses or events that occurred on or before index date.
TF, transfeminine; TM transmasculine, KPNC, Kaiser Permanente Northern California, KPGA, Kaiser Permanente Giorgia, KPSC, Kaiser Permanente Southern California.
Tables 3 and 4 show results for the multivariable Cox regression analyses. In the overall multivariable model for time to first event (Table 3), the hazard of attempting suicide at least once during follow-up did not differ between TM and TF cohort members (HR=0.95, 95% CI=0.73, 1.22). Relative to participants aged >45 years at baseline, the HR estimates were 3.26 (95% CI=1.77, 6.00) for those aged 36–45 years, 2.82 (95% CI=1.53, 5.20) for those aged 26– 35 years, 3.25 (95% CI=1.84, 5.75) for those aged 18–25 years, and 7.33 (95% CI=4.32, 12.43) for those aged <18 years. The HR estimate per each additional year when age was used as a continuous variable was 0.96 (95% CI=0.95, 0.97). In addition, relative to those with no mental health diagnosis, the HR estimates were 2.34 (95% CI=1.65, 3.32) for those with one to two diagnoses and 3.68 (95% CI=2.47, 5.48) for those with more than two diagnoses. The HR estimate for those with a history of suicidal ideation or previous suicide attempt at index date was 3.62 (95% CI=2.58, 5.08). (Table 3).
Table 3.
Multivariable Cox Proportional Hazards Regression Analysesa of Factors Associated With Incidence of the First Self-Harmb in the Cohort: Overall and by TF/TM Status (N=6,327)
| Participant characteristics | Overall cohort HR (95% CI) |
TF cohort HR (95% CI) |
TM cohort HR (95% CI) |
|---|---|---|---|
| TF/TM status | |||
| TF | 1.00 (ref) | NA | NA |
| TM | 0.95 (0.73, 1.22) | NA | NA |
| Age at index date, years | |||
| >45 | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| 36–45 | 3.26 (1.77, 6.00) | 4.83 (2.34, 9.94) | 1.56 (0.51, 4.80) |
| 26–35 | 2.82 (1.53, 5.20) | 3.65 (1.69, 7.88) | 2.09 (0.76, 5.71) |
| 18–25 | 3.25 (1.84, 5.75) | 3.76 (1.79, 7.89) | 2.57 (1.01, 6.55) |
| 3–17 | 7.33 (4.32, 12.43) | 6.84 (3.49, 13.41) | 6.70 (2.76, 16.28) |
| Number of mental health diagnoses at baselinec | |||
| None | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| 1–2 | 2.34 (1.65, 3.32) | 2.13 (1.34, 3.40) | 2.66 (1.55, 4.56) |
| >2 | 3.68 (2.47, 5.48) | 3.05 (1.78, 5.22) | 4.64 (2.53, 8.52) |
| History of suicidal ideation or self-harm at baselinec | |||
| No | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Yes | 3.62 (2.58, 5.08) | 4.61 (2.75, 7.74) | 3.01 (1.94, 4.69) |
Stratified on race/ethnicity and study site (KPNC vs Other).
Includes any diagnosis of self-inflicted injury, possibly self-inflicted injury, and self-inflicted injury/poisoning.
Baseline refers to diagnoses that occurred on or before index date.
TF, transfeminine; TM, transmasculine; HR, hazard ratio; NA, not applicable.
Table 4.
Counting Process Analysisa of Factors Associated With Incidence of Recurrent Self-Harmb in the Cohort: Overall and by TF/TM Status (N=6,327)
| Participant characteristics | Overall cohort HR (95% CI) |
TF cohort HR (95% CI) |
TM cohort HR (95% CI) |
|---|---|---|---|
| TF/TM status | |||
| TF | 1.00 (ref) | NA | NA |
| TM | 0.93 (0.70, 1.25) | NA | NA |
| Age at the end of last event, years | |||
| >45 | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| 36–45 | 3.00 (1.69, 5.32) | 4.53 (2.28, 9.02) | 1.34 (0.50, 3.65) |
| 26–35 | 3.75 (2.00, 7.03) | 3.76 (1.78, 7.95) | 3.40 (1.27, 9.12) |
| 18–25 | 2.87 (1.65, 4.97) | 3.57 (1.77, 7.21) | 2.13 (0.90, 5.03) |
| 3–17 | 6.31 (3.75, 10.62) | 6.00 (3.11, 11.59) | 5.66 (2.46, 12.98) |
| Number of mental health diagnoses at baselinec | |||
| None | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| 1–2 | 2.43 (1.68, 3.51) | 2.17 (1.36, 3.48) | 2.92 (1.64, 5.22) |
| >2 | 2.46 (1.60, 3.77) | 2.02 (1.14, 3.61) | 3.25 (1.74, 6.05) |
| History of suicidal ideation or self-harm at baselined | |||
| No | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Yes | 9.15 (6.52, 12.83) | 10.89 (6.91, 17.14) | 7.79 (4.95, 12.25) |
Stratified by race/ethnicity and study site (KPNC vs Other).
Includes any diagnosis of self-inflicted injury, possibly self-inflicted injury, and self-inflicted injury/poisoning.
Mental health diagnoses baseline refers to diagnoses that occurred on or before index date.
Suicidal ideation or self-harm baseline refers to events that occurred before each event.
TF, transfeminine; TM, transmasculine; HR, hazard ratio; NA, not applicable.
Table 4 presents the model that takes into consideration repeated events with age and history of suicide attempt or suicidal ideation measured at the start of each follow-up interval. In the cohort overall, younger patients were more likely to attempt suicide compared with those aged >45 years, with HR estimates of 3.00 (95% CI=1.69, 5.32) for those aged 36–45 years, 3.75 (95% CI=2.00, 7.03) for those aged 26–35 years, 2.87 (95% CI=1.65, 4.97) for those aged 18–25 years, and 6.31 (95% CI=3.75, 10.62) for those aged 3–17 years. The HR estimate for each additional year of age was 0.97 (95% CI=0.96, 0.98). Participants with a prior history of suicidal ideation or suicide attempt were significantly more likely to attempt suicide compared with those who had no such history at baseline (HR=9.15, 95% CI=6.52, 12.83).
DISCUSSION
This longitudinal study examined factors associated with incidence of attempted suicide in a cohort of more than 6,000 TGD individuals across three integrated healthcare systems. The incidence of suicide attempts among this large cohort of transgender patients was high, with 3.0% of TF and 4.8% of TM participants experiencing self-harm over the 10-year study period. Suicide attempts were especially common among individuals with a greater number of mental health diagnoses and those who had already experienced suicidal ideation or suicide attempts. The incidence also decreased with age.
Although comparisons are difficult owing to differing timeframes, only 0.2% of patients in a general clinical population presented to the emergency department annually under similar circumstances in a recent national study26—a much lower rate than the patient population described here. On the other hand, rates of suicide attempts that resulted in presentation to the emergency department among TGD youth in this study were 18–144 times higher among TGD youth compared to reference male and female youth.27 In addition, a recent study found that 18% of transgender veterans had documented suicidal ideation or attempts in their VHA records20—similar to the proportion of participants with suicidal ideation or attempted suicide prior to baseline found here. Though overall proportions of individuals with a suicide attempt history are substantially higher in previous studies compared with this one, this could be an artifact of study design. Unlike previous studies, which employed cross-sectional surveys, this study used longitudinal EMR data and examined factors associated with incidence rather than history of self-harm and related events. However, there may be other explanations for the lower suicide attempt rates found here. This study included people who may question their gender identity but experience no gender dysphoria and express no desire to receive gender-affirming care. In addition, though diagnostic codes were used for outcome ascertainment, the preponderance of prior studies relied on self-report and are not limited to episodes of self-harm that warrant an emergency department visit. The differences between population surveys and real world clinical data may also indicate that there is a need for improved identification and screening efforts with this population in clinical care settings with accompanying protocols for mental health referrals. TGD patients may not always go to the emergency department for suicide attempts, and better efforts are needed to engage this vulnerable population in care.
Earlier surveys and one-on-one interview studies that recruited convenience samples of TGD participants also found that younger TGD people may be at higher risk for suicide attempts compared with older TGD individuals.18,28,29 In addition, these age differences were found among respondents to the 2008 cross-sectional National Transgender Discrimination Survey; participants who were aged 45 years or younger had higher odds of ever attempting suicide compared with those who were older.28 This trend was also evident among respondents to the 2015 U.S. Transgender Survey: 10% of respondents aged 18–25 years reported attempting suicide in the past year, and these rates steadily decreased with increasing participant age.30 Although incidence of suicide attempts did not differ among TM and TF participants in this study, others who have conducted surveys have observed that TM individuals were at least slightly more likely to report such events.15,17,31
Self-inflicted injury diagnoses were used in this study as a surrogate for suicide attempts. Previous EMR-based studies addressing this issue in predominantly cisgender populations used more sophisticated algorithms that not only included data on self-inflicted injuries but also included combined data on suicidal ideation with information on any type of injury to more fully ascertain suicide attempts.32 Yet, data used here represent real-world suicide attempts as viewed and recorded by health professionals. This approach has been previously validated by other clinical studies as an accurate way to capture suicide attempts.22,23
The study timeframe (2006–2016) saw numerous sociopolitical changes and events that likely impacted the mental health of transgender Americans in both positive and negative ways. Positive strides for sexual and gender minorities (SGM) were made by the Obama administration during this time, including the signing of the Matthew Shepard and James Byrd, Jr. Hate Crimes Prevention Act (2009), the Don’t Ask, Don’t Tell Repeal Act (2010), and expansion of healthcare for many SGM Americans under the Patient Protection and Affordable Care Act (2010). Marriage equality, linked to more positive mental health among SGM individuals,33 was also on center stage politically during this time. In particular, the State of California was a prominent battle ground for marriage equality throughout the decade. California was also one of the first states to establish non-discrimination protections for TGD people, including requirements of private insurance coverage for gender-affirming care, while the rest of the country was seeing challenges to transgender protections in the form of bathroom bills.34,35 Although the State of California made progress in protecting transgender individuals, greater uncertainty was introduced nationally with the incoming Trump administration at the end of the study period, with SGM Americans reporting higher rates of minority stress, depression, and anxiety after the 2016 U.S. presidential election.36 Suicide attempts may have been comparatively lower among transgender California residents during this time, who experienced relatively better legal protections compared with those in other geographic regions. Future studies should examine how suicide attempt rates among gender minorities are impacted by local, state, and national political climate and protections.
Limitations
This study has several limitations. Although TGD people enrolled in integrated healthcare systems represent a cohort of individuals with health insurance that may not be representative of the TGD population in the U.S., this cohort does include patients enrolled in Medicaid plans, insuring at least some representation of patients with lower SES. In addition, the use of insured populations allows for better capture of both within- and outside-system use, which would not be possible to capture among uninsured populations. Moreover, the vast majority of the cohort members resided in California. It is expected that some of the results may differ among TGD people in different socioeconomic strata and geographic locations. Weighing against this concern is the demonstrated ability to cost effectively identify and follow a large cohort of TGD subjects with a high degree of internal validity. This study also includes both minors and adult patients. For individuals aged younger than 18 years, emergency department visits and access are likely driven by parents or other caregivers, so comparing emergency department visits for self-harm across age groups should be done cautiously. Also, subcategorization of patients aged older than 45 years was not possible, which limits the ability to draw conclusions about older age groups. In addition, TGD people who receive appropriate support and gender-affirming care may experience improvement in their mental health status and overall quality of life.37–41 It follows that people who achieve greater congruence between their gender identity and appearance may also experience a reduction in suicidal ideation and suicide attempts. Future studies should explore whether and how receipt of appropriate gender-affirming care influences suicide risk. Finally, suicide deaths were not examined as part of this study. Others have found that among transgender veterans, rates of suicide deaths are higher than those of cisgender veterans as well as the U.S. population,42 and further research using EMR data is needed to determine the rate of suicide deaths among transgender individuals in the U.S.
CONCLUSIONS
The most powerful predictors of suicide attempts among TGD individuals are mental health diagnoses at baseline and past history of suicidal ideation or suicide attempts. Younger TGD people may also be at higher risk for attempted suicide compared with their older counterparts independently of other demographic characteristics or mental health status. Although these predictors also hold true for the general population,43 what is unclear is the extent to which risk factors such as mental health diagnoses explain disparities in suicide attempt rates versus factors that are unique to TGD individuals. To that end, future studies should investigate the impact of gender-affirming care and the influence of specific mental health diagnoses on suicide risk among this population. These data will be important for risk stratification and for the development and implementation of interventions aimed at preventing suicide among TGD people.
ACKNOWLEDGMENTS
Funding sources for this work included Contract AD-12-11-4532 from the Patient Centered Outcome Research Institute and Grant R21HD076387 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Previous Presentation: Society for Social Work and Research (2019), San Francisco, CA.
Dr. Flanders provides consulting services through his company Epidemiologic Research & Methods, LLC. He knows of no conflicts. No other financial disclosures were reported by the authors of this paper.
REFERENCES
- 1.Mayer KH, Bradford JB, Makadon HJ, Stall R, Goldhammer H, Landers S. Sexual and gender minority health: what we know and what needs to be done. Am J Public Health. 2008;98(6):989–995. 10.2105/ajph.2007.127811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Graham R, Berkowitz B, Blum R, et al. The health of lesbian, gay, bisexual, and transgender people: building a foundation for better understanding. Washington, DC: Institute of Medicine; 2011:89–139. 10.17226/13128. [DOI] [PubMed] [Google Scholar]
- 3.Conron KJ, Scott G, Stowell GS, Landers SJ. Transgender health in Massachusetts: results from a household probability sample of adults. Am J Public Health. 2012;102(1):118–122. 10.2105/ajph.2011.300315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Crissman HP, Berger MB, Graham LF, Dalton VK. Transgender demographics: a household probability sample of US adults, 2014. Am J Public Health. 2017;107(2):213–215. 10.2105/ajph.2016.303571. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Meerwijk EL, Sevelius JM. Transgender population size in the United States: a meta-regression of population-based probability samples. Am J Public Health. 2017;107(2):216 10.2105/ajph.2016.303578a. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bockting WO, Miner MH, Swinburne Romine RE, Hamilton A, Coleman E. Stigma, mental health, and resilience in an online sample of the US transgender population. Am J Public Health. 2013;103(5):943–951. 10.2105/ajph.2013.301241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Brown GR, Jones KT. Mental health and medical health disparities in 5135 transgender veterans receiving healthcare in the Veterans Health Administration: a case-control study. LGBT Health. 2016;3(2):122–131. 10.1089/lgbt.2015.0058. [DOI] [PubMed] [Google Scholar]
- 8.Reisner SL, Vetters R, Leclerc M, et al. Mental health of transgender youth in care at an adolescent urban community health center: a matched retrospective cohort study. J Adolesc Health. 2015;56(3):274–279. 10.1016/j.jadohealth.2014.10.264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Clements-Nolle K, Marx R, Guzman R, Katz M. HIV prevalence, risk behaviors, health care use, and mental health status of transgender persons: implications for public health intervention. Am J Public Health. 2001;91(6):915–921. 10.2105/ajph.91.6.915. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Blosnich JR, Brown GR, Shipherd JC, Kauth M, Piegari Rl, Bossarte RM. Prevalence of gender identity disorder and suicide risk among transgender veterans utilizing Veterans Health Administration care. Am J Public Health. 2013;103(10):e27–e32. 10.2105/ajph.2013.301507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Clements-Nolle K, Marx R, Katz M. Attempted suicide among transgender persons: the influence of gender-based discrimination and victimization. J Homosex. 2006;51(3):53–69. 10.1300/j082v51n0304. [DOI] [PubMed] [Google Scholar]
- 12.Haas AP, Eliason M, Mays VM, et al. Suicide and suicide risk in lesbian, gay, bisexual, and transgender populations: review and recommendations. J Homosex. 2011;58(1):10–51. 10.1080/00918369.2011.534038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Skerrett DM, Kolves K, De Leo D. Suicides among lesbian, gay, bisexual, and transgender populations in Australia: an analysis of the Queensland Suicide Register. Asia Pac Psychiatry. 2014;6(4):440–446. 10.1111/appy.12128. [DOI] [PubMed] [Google Scholar]
- 14.Nock MK, Borges G, Bromet EJ, Cha CB, Kessler RC, Lee S. Suicide and suicidal behavior. Epidemiol Rev. 2008;30(1):133–154. 10.1093/epirev/mxn002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Herman JL, Haas AP, Rodgers PL. Suicide attempts among transgender and gender nonconforming adults. https://escholarship.org/uc/item/8xg8061f. Published 2014. Accessed April 8, 2020. [Google Scholar]
- 16.Grant JM, Mottet LA, Tanis J, Herman JL, Harrison J, Keisling M. National transgender discrimination survey report on health and health care. Washington, DC: National Center for Transgender Equality and the National Gay and Lesbian Task Force; 2010. [Google Scholar]
- 17.Perez-Brumer A, Hatzenbuehler ML, Oldenburg CE, Bockting W. Individual- and structural-level risk factors for suicide attempts among transgender adults. Behav Med. 2015;41(3):164–171. 10.1080/08964289.2015.1028322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Clements-Nolle K, Marx R, Katz M. Attempted suicide among transgender persons: the influence of gender-based discrimination and victimization. J Homosex. 2006;51(3):53–69. 10.1300/j082v51n0304. [DOI] [PubMed] [Google Scholar]
- 19.Mustanski B, Liu RT. A longitudinal study of predictors of suicide attempts among lesbian, gay, bisexual, and transgender youth. Arch Sex Behav. 2013;42(3):437–448. 10.1007/sl0508-012-0013-9. [DOI] [PubMed] [Google Scholar]
- 20.Blosnich JR, Marsiglio MC, Dichter ME, et al. Impact of social determinants of health on medical conditions among transgender veterans. Am J Prev Med. 2017;52(4):491–498. 10.1016/j.amepre.2016.12.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Quinn VP, Nash R, Hunkeler E, et al. Cohort profile: Study of Transition, Outcomes and Gender (STRONG) to assess health status of transgender people. BMJ Open. 2017;7(12):e018121 10.1136/bmjopen-2017-018121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Simon GE, Coleman KJ, Rossom RC, et al. Risk of suicide attempt and suicide death following completion of the Patient Health Questionnaire depression module in community practice. J Clin Psychiatry. 2016;77(2):221–227. 10.4088/jcp.15m09776. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Rossom RC, Coleman KJ, Ahmedani BK, et al. Suicidal ideation reported on the PHQ9 and risk of suicidal behavior across age groups. J Affect Disord. 2017;215:77–84. 10.1016/j.jad.2017.03.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Silverberg MJ, Nash R, Becerra-Culqui TA, et al. Cohort study of cancer risk among insured transgender people. Ann Epidemiol. 2017;27(8):499–501. 10.1016/j.annepidem.2017.07.007. [DOI] [PubMed] [Google Scholar]
- 25.Amorim LD, Cai J. Modelling recurrent events: a tutorial for analysis in epidemiology. Int J Epidemiol. 2015;44(l):324–333. 10.1093/ije/dyu222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Canner J, Giuliano K, Selvarajah S, Hammond E, Schneider E. Emergency department visits for attempted suicide and self harm in the USA: 2006-2013. Epidemiol Psychiatr Sci. 2018;27(1):94–102. 10.1017/s2045796016000871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Becerra-Culqui TA, Liu Y, Nash R, et al. Mental health of transgender and gender nonconforming youth compared with their peers. Pediatrics. 2018;141(5):e20173845 10.1542/peds.2017-3845. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Klein A, Golub SA. Family rejection as a predictor of suicide attempts and substance misuse among transgender and gender nonconforming adults. LGBT Health. 2016;3(3):193–199. 10.1089/lgbt.2015.0111. [DOI] [PubMed] [Google Scholar]
- 29.Goldblum P, Testa RJ, Pflum S, Hendricks ML, Bradford J, Bongar B. The relationship between gender-based victimization and suicide attempts in transgender people. Prof Psychol Res Pr. 2012;43(5):468–475. 10.1037/a0029605. [DOI] [Google Scholar]
- 30.James SE, Herman JL, Rankin S, Keisling M, Mottet LA, Anafi M. The Report of the 2015 U.S. Transgender Survey. https://transequalitv.org/sites/default/files/docs/usts/USTS-Full-Report-Decl7.pdf. Published 2016. Accessed April 8, 2020. [Google Scholar]
- 31.Maguen S, Shipherd JC. Suicide risk among transgender individuals. Psychol Sex. 2010;1(1):34–43. 10.1080/19419891003634430. [DOI] [Google Scholar]
- 32.Simon GE, Johnson E, Lawrence JM, et al. Predicting suicide attempts and suicide deaths following outpatient visits using electronic health records. Am J Psychiatry. 2018;175(10):951–960. 10.1176/appi.ajp.2018.17101167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Perone AK. Health implications of the Supreme Court’s Obergefell vs. Hodges marriage equality decision. LGBT Health. 2015;2(3):196–199. 10.1089/lgbt.2015.0083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Schilt K, Westbrook L. Bathroom battlegrounds and penis panics. Contexts. 2015;14(3):26–31. 10.1177/1536504215596943. [DOI] [Google Scholar]
- 35.Lee C. The fight for children’s ‘safety’in North Carolina. The Brown University Child and Adolescent Behavior Letter. 2016;32(6):8 10.1002/cbl.30133. [DOI] [Google Scholar]
- 36.Gonzalez KA, Ramirez JL, Galupo MP. Increase in GLBTQ minority stress following the 2016 US presidential election. J GLBT Fam Stud. 2018;14(1–2):130–151. 10.1080/1550428x.2017.1420849. [DOI] [Google Scholar]
- 37.White Hughto JM, Reisner SL. A systematic review of the effects of hormone therapy on psychological functioning and quality of life in transgender individuals. Transgend Health. 2016;1(1):21–31. 10.1089/trgh.2015.0008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Pauly IB, Lindgren TW. Body image and gender identity. J Homosex. 1976;2(2):133–142. 10.1300/j082v02n0204. [DOI] [PubMed] [Google Scholar]
- 39.Owen-Smith AA, Gerth J, Sineath RC, et al. Association between gender confirmation treatments and perceived gender congruence, body image satisfaction, and mental health in a cohort of transgender individuals. J Sex Med. 2018;15(4):591–600. 10.1016/j.jsxm.2018.01.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Fisher AD, Castellini G, Ristori J, et al. Cross-sex hormone treatment and psychobiological changes in transsexual persons: two-year follow-up data. J Clin Endocrinol Metab. 2016; 101(11):4260–4269. 10.1210/jc.2016-1276. [DOI] [PubMed] [Google Scholar]
- 41.Lindgren TW, Pauly IB. A body image scale for evaluating transsexuals. Arch Sex Behav. 1975;4(6):639–656. 10.1007/bf01544272. [DOI] [PubMed] [Google Scholar]
- 42.Blosnich JR, Brown GR, Wojcio S, Jones KT, Bossarte RM. Mortality among veterans with transgender-related diagnoses in the Veterans Health Administration, FY2000–2009. LGBT Health. 2014; 1(4):269–276. 10.1089/lgbt.2014.0050. [DOI] [PubMed] [Google Scholar]
- 43.Office of the Surgeon General (US), National Action Alliance for Suicide Prevention (US). 2012 national strategy for suicide prevention: goals and objectives for action: a report of the US Surgeon General and of the National Action Alliance for Suicide Prevention. Washington, DC: HHS; 2012. [PubMed] [Google Scholar]