Abstract
Purpose: Transgender and nonbinary people have an increased burden of psychiatric problems compared with the general population. Data are needed to understand factors associated with psychiatric diagnoses, acuity in terms of suicide attempts and level-of-care escalation, and outpatient engagement among transgender and nonbinary adults.
Methods: We conducted a retrospective review of records from 201 transgender and nonbinary adults who presented for primary care at a health center. Regression models were fit to examine factors associated with psychiatric diagnoses, substance use disorders (SUDs), acuity, and outpatient behavioral health engagement.
Results: Male sex assignment at birth was associated with decreased odds of a psychiatric diagnosis (odds ratio [OR] 0.40, 95% confidence interval [CI]: 0.20–0.81). Increased odds of SUDs were associated with later hormone initiation (OR 1.04, 95% CI: 1.01–1.08) and suicide attempt (OR 5.79, 95% CI: 2.08–16.15). Increased odds of higher acuity were associated with alcohol use disorder (OR 31.54, 95% CI: 5.73–173.51), post-traumatic stress disorder (OR 18.14, 95% CI: 2.62–125.71), major depressive disorder (MDD) (OR 6.62, 95% CI: 1.72–25.44), and absence of psychiatrist integration into primary medical care (OR 4.52, 95% CI: 1.26–16.22). Increased odds of outpatient behavioral health engagement were associated with case management utilization (OR 10.73, 95% CI: 1.32–87.53), anxiety disorders (OR 15.84, 95% CI: 2.00–125.72), and MDD (OR 10.45, 95% CI: 2.28–47.98).
Conclusion: Psychiatric disorders were highly prevalent among transgender and nonbinary adult patients. Novel findings include associations of lack of psychiatrist integration into primary care with acuity and of case management utilization with outpatient behavioral health engagement.
Keywords: mental health, minority stress, nonbinary, substance use disorders, suicide, transgender
Introduction
Atransgender person has a gender identity that differs from the gender traditionally associated with their sex assigned at birth; a cisgender person has a gender identity that does not differ from the gender typically associated with their sex assigned at birth.1,2 Approximately 1.4 million adults in the United States identify as transgender.3 Gender identity can be binary—that is, anchored to a conventional girl/woman-boy/man binary paradigm—or nonbinary—that is, outside girl/woman or boy/man categorization. Nonbinary people identify as neither girl/woman nor boy/man.1,4 Among gender minority people, lived experiences and health needs differ substantially across the continuum of nonbinary and binary gender identities, as well as based on being assigned female sex at birth (AFAB) versus assigned male sex at birth (AMAB).5,6
Although studies elucidating mental health outcomes among transgender and nonbinary people are beginning to emerge, further research is needed to characterize many aspects of psychiatric epidemiology among gender minority people.7–16 Prior findings suggest that transgender people have increased psychiatric problems compared with the general population.7,13–17 A 2013 analysis of cross-sectional data from both trans masculine and trans feminine people in a large, diverse online sample found that respondents had a high prevalence of clinical depression (44.1%) and anxiety (33.2%).13 Secondary data analysis of a 2013 community-based survey in Massachusetts exploring substance use disorders (SUDs) among transgender adults found that 10% reported lifetime SUD treatment,7 compared with 8.4% of the general population.18 A 2017 systematic review of studies investigating suicidality in transgender people found prevalences of suicidal ideation ranging from 37% to 83%.19 Moreover, the 2015 U.S. Transgender Survey documented a 40% lifetime prevalence of suicide attempts among transgender and nonbinary people compared with 4.6% in the general population.20
One possible explanation for these disparities is the potential role of gender minority stress in precipitating psychopathology. Chronic stress associated with stigma regarding minority status can be internalized and embodied, potentially increasing the likelihood of psychiatric disorders.21 It is known that transgender and nonbinary people experience higher levels of stigmatization than cisgender people,6 and these experiences are believed to be associated with higher likelihood of onset of some psychiatric disorders in particular, such as major depressive disorder (MDD), anxiety disorders,5 post-traumatic stress disorder (PTSD),22,23 and SUDs. This is in contrast to diagnoses for which there is less evidence of onset secondary to chronic gender minority stress, such as bipolar affective disorder (BPAD) and personality disorders.24–27
Studies of psychiatric treatment utilization among transgender people are sparse.28–32 In one San Francisco-based study, despite higher prevalence of physical and mental health problems and psychopharmacology use among transgender people compared with cisgender people, there was no difference between transgender and cisgender people in emergency room visits, outpatient emergency behavioral health care utilization, or psychiatric hospitalizations.28 Although this finding could reflect more appropriate utilization of nonemergent care, a plausible and concerning alternate conclusion is that transgender patients are not accessing care when they need it, perhaps due to fears of misgendering and mistreatment. A consistent finding is that culturally affirming behavioral health services for transgender people are lacking.29 Further study of psychiatric care utilization is critical to improve mental health outcomes among gender minority people.
The aims of this study were to assess psychiatric diagnoses (with SUDs examined separately), acuity in terms of suicide attempts and level-of-care escalation, and outpatient behavioral health engagement among transgender and nonbinary adults, and to investigate associations with demographics, gender identity-related characteristics, sexual history, and other mental health factors. We hypothesized that (1) factors specifically linked with minority stress would be associated with higher odds of psychiatric diagnoses and (2) these factors, as well as psychiatric diagnoses, which are more likely to develop from internalizing minority stress-related stigma, would be associated with higher odds of psychiatric acuity among transgender and nonbinary adults.7–16,33,34
Methods
Study participants and procedures
We undertook a retrospective chart review of electronic health record (EHR) data from a random sample of 201 transgender and nonbinary patients 18–64 years of age, who sought at least one primary medical care visit between July 1, 2010, and June 30, 2015, at a U.S. federally qualified health center (FQHC) specializing in lesbian, gay, bisexual, transgender, and queer (LGBTQ) care.35,36 Patients were identified as transgender or nonbinary based on a standardized two-step EHR designation system, where sex assigned at birth (coded as binary) and current gender identity (“male,” “female,” “trans masculine,” “trans feminine,” or “neither exclusively male nor female”) are recorded sequentially. Based on this two-step system, a patient would be identified as transgender, for example, if they indicated they were AMAB and currently identified as female. The study sample (N = 201) was selected using an automated simple random sampling algorithm from all transgender and nonbinary patients 18 years or older (N = 1683).37
Demographic variables, sex assigned at birth, gender identity, sexual history, and mental health factors were extracted from the EHR, using both automated query and manual chart audit, and included as statistical predictor variables. To minimize recall error, missing data, and other biases, variables were operationally defined by specific parameters and collected by prespecified systematic protocols.38,39 The study was approved by the Institutional Review Board at Fenway Health. All patients had been informed, in keeping with the FQHC's privacy policy, that their EHR information may be used for research purposes.
Binary outcome variables
We constructed four outcome variables. We defined the first outcome variable based on the presence of ≥1 of the following: PTSD, any anxiety disorder, MDD, BPAD, and/or any personality disorder. These disorders were grouped to include a diversity of psychiatric diagnoses that are commonly seen and treated in outpatient community psychiatry settings, and because they can relatively consistently be extracted from the FQHC's EHR data. To explore factors specifically associated with addictions, SUDs were separated out as a second outcome variable (any current or past alcohol, cannabis, cocaine, opioid, benzodiazepine, amphetamine, or intravenous drug use disorder). Diagnoses were initially obtained by automated query of EHR problem lists that feature the diagnoses entered by the patients' psychiatrists, primary care physicians, or psychotherapists at the FQHC. Thereafter, one author who is a practicing psychiatrist at the FQHC conducted manual EHR audit of documentation from diagnostic clinical encounters to confirm the validity of these diagnoses using Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) diagnostic criteria.40
The third outcome variable was psychiatric acuity, defined as a history of ≥1 of the following: any past suicide attempt, inpatient hospitalization, partial hospitalization program, or residential treatment program. We grouped these acuity-related outcomes in light of the clinical significance inherent in any escalation of care beyond the outpatient level. The fourth outcome variable was outpatient behavioral health engagement, defined as engagement with ≥1 of the following: any psychopharmacology, psychotherapy, and/or SUD visits.
Statistical predictors
We assessed four categories of statistical predictors of each outcome variable. The statistical predictor variable categories included demographics (age, race/ethnicity, employment, and income), gender identity-related characteristics (nonbinary vs. binary identity [given known differences in minority stress and psychiatric epidemiology],6 gender-affirming hormone therapy, gender-affirming surgery, and age at hormone therapy initiation), sexual history [sexual orientation, presence vs. absence of casual sex partner(s), presence vs. absence of primary sex partner(s), and any past sexually transmitted infection(s)], and other mental health factors. This last category included psychiatric diagnoses, substance use and SUDs, indicators of higher psychiatric acuity, outpatient behavioral health engagement, and case management utilization. Many of these variables are included in outcome variables defined in the previous section; a variable was only included as a predictor if it was not part of the outcome for the particular analysis being conducted. We included these statistical predictor variables because we hypothesized, based on clinical experience and previous data, that they may be associated with psychiatric health outcomes.41–43 These characteristics were ascertained from the EHR as described in the previous section.
Statistical analysis
Univariate statistics were used to examine the variable distributions overall and stratified by assigned sex at birth. Bivariate analyses were conducted to compare AFAB and AMAB patients. Mann–Whitney U tests were used to assess median differences for non-normally distributed continuous variables (i.e., “current age” and “age at hormone therapy initiation”). Pearson's chi-square tests with Yates' correction were used to examine any difference in expected and observed proportions by gender identity (binary vs. nonbinary). Where sparse data caused expected counts to be <5, Fisher's exact tests were utilized to obtain p-values. Some of these univariate and bivariate statistics were previously published in a study with the same patient sample that instead investigated factors associated with gender-affirming surgery and age at hormone therapy initiation.44
Multivariable logistic regression analyses were conducted with variables that had >85% completeness. Fifty-six patients (27.9%) who did not have complete data for all variables were excluded from multivariable regression procedures; analyses were thus restricted to 145 patients (72.1% of the original 201-person sample; 73 AFAB and 72 AMAB). Chi-square analyses demonstrated no statistically significant differences in the distributions of assigned sex or gender identity between the included and excluded groups. Rather than conducting analyses with data obtained by automated EHR query about the health center's overall pool of transgender and nonbinary adults, we prioritized conducting rigorous manual EHR audit of clinical notes for confirmation of variables assessed from a smaller random sample of patients, a labor-intensive process to increase the validity of patient-related information obtained in the EHR review. We therefore conducted subsequent chi-square analyses comparing the analytic group (N = 145) with the health center's overall pool of 1638 transgender and nonbinary patients across each race/ethnicity category and found no statistically significant differences. A dummy variable coding patients excluded from the analytic sample in multivariable analyses was created and analyzed as a bivariate against each regression outcome using the full study sample (N = 201) to assess whether patient exclusion introduced bias.
To increase statistical power, the sample was analyzed in aggregate and not stratified by gender identity or sex assigned at birth; these variables were included as predictor variables. Model building initially focused on bivariate models for variables with statistical significance. Factors significant at the p < 0.05 level were included in a multivariable model. Backward elimination was utilized to select for independent associations such that variables were sequentially eliminated from the model using the smallest chi-square-to-remove at a threshold for inclusion of p < 0.05.45 Data analyses were conducted using SAS Studio Release 3.5, 2016 (SAS Institute Inc., Cary, NC), and Microsoft Excel 2010 (Microsoft Corporation, Redmond, WA).
Results
Descriptive and bivariate analyses
Analytic sample demographics, gender identity-related characteristics, sexual history, and other mental health factors are presented in Tables 1–4 for AFAB and AMAB participants separately and in aggregate. By chance, half of the sample was AFAB and the other half was AMAB, reflecting the overall composition of the transgender and nonbinary populations at the FQHC. Table 5 provides descriptive statistics about co-occurrence of psychiatric diagnoses and SUDs in the sample.
Table 1.
Demographics of the Analytic Sample
| Variable | AFAB (N = 73) | AMAB (N = 72) | Total (N = 145) | AFAB vs. AMAB, p |
|---|---|---|---|---|
| Age in years | ||||
| Mean (SD) | 27.9 (6.9) | 35.7 (13.7) | 31.8 (11.5) | |
| Median | 25.0 | 30.0 | 27.0 | 0.001 |
| Range | 19–50 | 21–64 | 19–64 | |
| Population age strata (years), N (%) | ||||
| 18–25 | 40 (54.8) | 24 (33.3) | 64 (44.1) | <0.001 |
| 26–49 | 32 (43.8) | 33 (45.8) | 65 (44.8) | |
| 50+ | 1 (1.4) | 15 (20.8) | 16 (11.0) | |
| Race/ethnicity | ||||
| White | 55 (75.3) | 60 (83.3) | 115 (79.3) | 0.235 |
| Black/African American | 5 (6.8) | 1 (1.4) | 6 (4.1) | |
| Latinx/Hispanic | 2 (2.7) | 3 (4.2) | 5 (3.4) | |
| Multiracial | 7 (9.6) | 5 (6.9) | 12 (8.3) | |
| Other | 3 (4.1) | 2 (2.8) | 5 (3.4) | |
| Not indicated | 1 (1.4) | 1 (1.4) | 2 (1.4) | |
| Employment | ||||
| Working full or part time | 51 (69.9) | 48 (66.7) | 99 (68.3) | 0.200 |
| Not working (unemployed, retired, or disabled) | 10 (13.7) | 17 (23.6) | 27 (18.6) | |
| Student | 12 (16.4) | 7 (9.7) | 19 (13.1) | |
| Income | 0.587 | |||
| At or below poverty level | 29 (39.7) | 31 (43.1) | 60 (41.4) | |
| 100%–200% of poverty level | 8 (11.0) | 11 (15.3) | 19 (13.1) | |
| 200%–300% of poverty level | 17 (23.3) | 11 (15.3) | 28 (19.3) | |
| Over 300% of poverty level | 17 (23.3) | 19 (26.4) | 36 (24.8) | |
| Not indicated | 2 (2.7) | 0 (0.0) | 2 (1.4) | |
Bold indicates statistical significance (p < 0.05). The response rate was 100% for all variables, except where the table states “Not indicated.”
AFAB, assigned female sex at birth; AMAB, assigned male sex at birth; SD, standard deviation.
Table 2.
Gender Identity-Related Characteristics of the Analytic Sample
| Variable | AFAB (N = 73), N (%) | AMAB (N = 72), N (%) | Total (N = 145), N (%) | AFAB vs. AMAB, p |
|---|---|---|---|---|
| Gender identity | 0.733 | |||
| Nonbinary | 19 (26.0) | 16 (22.2) | 35 (24.1) | |
| Binary | 54 (74.0) | 56 (77.8) | 110 (75.9) | |
| Hormones prescribed by primary care provider | 69 (94.5) | 69 (95.8) | 138 (95.2) | 1.000 |
| Current medically unmonitored hormone use | 0 (0.0) | 3 (4.2) | 3 (2.1) | 0.120 |
| Any past medically unmonitored hormone use | 0 (0.0) | 6 (8.3) | 6 (4.1) | 0.013 |
| Any gender-affirming surgery | 27 (37.0) | 20 (27.8) | 47 (32.4) | 0.314 |
| Age at hormone therapy initiation | ||||
| Mean (SD) | 27.9 (7.1) | 33.3 (13.2) | 31.8 (11.1) | |
| Median | 24.0 | 27.0 | 27.0 | 0.001 |
| Range | 8–50 | 15–64 | 8–64 | |
Bold indicates statistical significance (p < 0.05). The response rate was 100% for all variables.
Table 3.
Sexual Orientation and History of the Analytic Sample
| Variable | AFAB (N = 73), N (%) | AMAB (N = 72), N (%) | Total (N = 145), N (%) | AFAB vs. AMAB, p |
|---|---|---|---|---|
| Sexual orientation | 0.019 | |||
| Bisexual | 9 (12.3) | 22 (30.6) | 31 (21.4) | |
| Lesbian, gay, or homosexual | 14 (19.2) | 15 (20.8) | 29 (20.0) | |
| Straight or heterosexual | 15 (20.5) | 9 (12.5) | 24 (16.6) | |
| Something else | 31 (42.5) | 18 (25.0) | 49 (33.8) | |
| Does not know | 4 (5.5) | 8 (11.1) | 12 (8.3) | |
| Primary sex partner(s) | 48 (65.8) | 42 (58.3) | 90 (62.1) | 0.454 |
| Casual sex partner(s) | 5 (6.8) | 5 (6.9) | 10 (6.9) | 0.982 |
| Any sexually transmitted infection | 13 (17.8) | 7 (9.7) | 20 (13.8) | 0.242 |
Bold indicates statistical significance (p < 0.05). The response rate was 100% for all variables.
Table 4.
Other Mental Health Factors for the Analytic Sample
| Variable | AFAB (N = 73), N (%) | AMAB (N = 72), N (%) | Total (N = 145), N (%) | AFAB vs. AMAB, p |
|---|---|---|---|---|
| Outcome: any assessed psychiatric diagnosis | 43 (58.9) | 30 (41.7) | 73 (50.3) | 0.056 |
| Post-traumatic stress disorder | 6 (8.2) | 3 (4.2) | 9 (6.2) | 0.494 |
| Anxiety disorder | 28 (38.4) | 13 (18.1) | 41 (28.3) | 0.011 |
| MDD | 30 (41.1) | 25 (34.7) | 55 (37.9) | 0.536 |
| Bipolar disorder | 2 (2.7) | 2 (2.8) | 4 (2.8) | 1.000 |
| Personality disorder | 2 (2.7) | 1 (1.4) | 3 (2.1) | 1.000 |
| Outcome: lifetime SUD | 13 (17.8) | 17 (23.6) | 30 (20.7) | 0.511 |
| Current alcohol use disorder | 3 (4.1) | 8 (11.1) | 11 (7.6) | 0.129 |
| Past alcohol use disorder | 5 (6.8) | 5 (6.9) | 10 (6.9) | 0.982 |
| Current cannabis use disorder | 7 (9.6) | 5 (6.9) | 12 (8.3) | 0.782 |
| Outcome: indicators of higher psychiatric acuity | 13 (17.8) | 11 (15.3) | 24 (16.6) | 0.852 |
| History of suicide attempt | 11 (15.1) | 9 (12.5) | 20 (13.8) | 0.836 |
| History of inpatient psychiatric treatment | 11 (15.1) | 6 (8.3) | 17 (11.7) | 0.316 |
| History of residential or partial hospitalization program | 2 (2.7) | 3 (4.2) | 5 (3.4) | 0.681 |
| Outcome: outpatient behavioral health engagement | 59 (80.8) | 47 (65.3) | 106 (73.1) | 0.054 |
| Current psychotherapist | 47 (64.4) | 42 (58.3) | 89 (61.4) | 0.564 |
| Current psychopharmacologist | 36 (49.3) | 29 (40.3) | 65 (44.8) | 0.354 |
| Psychiatrist integrated with primary care | 3 (4.1) | 3 (4.2) | 6 (4.1) | 1.000 |
| Addictions program integrated with primary care | 0 (0.0) | 2 (2.8) | 2 (1.4) | 0.245 |
| Psychiatrist elsewhere | 22 (30.1) | 13 (18.1) | 35 (24.1) | 0.132 |
| Lifetime substance use | 59 (80.8) | 54 (75.0) | 113 (77.9) | 0.519 |
| Current alcohol use | 52 (71.2) | 49 (68.1) | 101 (69.7) | 0.814 |
| Past alcohol use | 7 (9.6) | 6 (8.3) | 13 (9.0) | 0.791 |
| Current cannabis use | 26 (35.6) | 22 (30.6) | 48 (33.1) | 0.638 |
| Current case management utilization | 18 (24.7) | 15 (20.8) | 33 (22.8) | 0.723 |
Bold indicates statistical significance (p < 0.05). The response rate was 100% for all variables.
MDD, major depressive disorder; SUD, substance use disorder.
Table 5.
Co-Occurrence of Psychiatric Diagnoses and Substance Use Disorders
| Variable | AFAB (N = 73), N (%) | AMAB (N = 72), N (%) | Total (N = 145), N (%) | AFAB vs. AMAB, p |
|---|---|---|---|---|
| Lifetime psychiatric diagnoses (excluding SUDs) | ||||
| 0 | 30 (41.1) | 42 (58.3) | 72 (49.7) | 0.056 |
| 1 | 23 (31.5) | 18 (25.0) | 41 (28.3) | 0.493 |
| 2 | 15 (20.5) | 10 (13.9) | 25 (17.2) | 0.400 |
| 3 | 5 (6.8) | 2 (2.8) | 7 (4.8) | 0.442 |
| Lifetime SUDs | ||||
| Total | 13 (17.8) | 17 (23.6) | 30 (20.7) | 0.511 |
| SUDs co-occurring with other psychiatric disorders | 11 (15.1) | 9 (12.5) | 20 (13.8) | 0.836 |
Multivariable regression models
Tables 6–9 present multivariable logistic regression models that examine factors associated with psychiatric diagnoses, SUDs, psychiatric acuity, and outpatient behavioral health engagement. Factors associated with a statistically significant increase in the probability of a non-SUD psychiatric diagnosis were AFAB and not working (Table 6). For this analysis alone, the dummy variable for excluded individuals was a significant bivariate statistical predictor of the outcome variable (p = 0.015). Inclusion in the analysis (based on having complete data for all desired variables) was significantly associated with lower odds of having a psychiatric diagnosis. Factors associated with increased odds of SUD were past suicide attempt and older age at hormone therapy initiation (Table 7).44
Table 6.
Significant Outcomes of Logistic Regression Models on Presence of Any Psychiatric Diagnosis (Excluding Substance Use Disorders)
| Variable | Bivariate models | Multivariable model | ||
|---|---|---|---|---|
| OR (95% CI) | p | OR (95% CI) | p | |
| Demographics | ||||
| AMAB | 0.50 (0.26–0.97) | 0.039 | 0.40 (0.20–0.81) | 0.011 |
| Not working (unemployed, retired, or disabled) | 4.42 (1.67–11.81) | 0.003 | 5.54 (2.00–15.33) | 0.001 |
| Dummy variable (tested with N = 201) | 0.44 (0.23–0.85) | 0.015 | ||
N = 145. Bold indicates statistical significance (p < 0.05). Mental health treatment utilization variables, including inpatient psychiatric treatment, residential or partial hospitalization program, and outpatient behavioral health engagement, were excluded from this regression to avoid using treatment modality variables to predict diagnoses.
CI, confidence interval; OR, odds ratio.
Table 7.
Significant Outcomes of Logistic Regression Models on Lifetime History of Substance Use Disorder
| Variable | Bivariate models | Multivariable model | ||
|---|---|---|---|---|
| OR (95% CI) | p | OR (95% CI) | p | |
| Gender identity-related characteristics | ||||
| Age at hormone therapy initiation | 1.04 (1.00–1.07) | 0.041 | 1.04 (1.01–1.08) | 0.025 |
| Mental health | ||||
| History of suicide attempt | 5.25 (1.93–14.25) | 0.001 | 5.79 (2.08–16.15) | <0.001 |
N = 145. Bold indicates statistical significance (p < 0.05). Mental health treatment utilization variables, including inpatient psychiatric treatment, residential or partial hospitalization program, and outpatient behavioral health engagement, were excluded from this regression to avoid using treatment modality variables to predict diagnoses.
Factors associated with a statistically significant increase in the odds of psychiatric acuity included current alcohol use disorder, PTSD, and MDD (Table 8). Personality disorders and BPAD were not associated with increased acuity in this patient population. An absence of integration of one's psychiatrist into primary medical care was significantly associated with increased odds of acuity, whereas having a psychiatrist integrated with primary medical care was not associated with acuity. There was a statistically significant association between outpatient behavioral health engagement and the following: case management utilization, any anxiety disorder, and MDD (Table 9). Other psychiatric statistical predictor variables, such as BPAD and personality disorders, were not associated with outpatient behavioral health engagement.
Table 8.
Significant Outcomes of Logistic Regression Models on Indicators of Higher Psychiatric Acuity
| Variable | Bivariate models | Multivariable model | ||
|---|---|---|---|---|
| OR (95% CI) | p | OR (95% CI) | p | |
| Demographics | ||||
| Not working (unemployed, retired, or disabled) | 4.37 (1.67–11.41) | 0.003 | ||
| Mental health | ||||
| Past alcohol use | 3.72 (1.10–12.57) | 0.035 | ||
| Lifetime SUD | 5.72 (2.23–14.71) | <0.001 | ||
| Current alcohol use disorder | 19.67 (4.73–81.85) | <0.001 | 31.54 (5.73–173.51) | <0.001 |
| Post-traumatic stress disorder | 24.50 (4.70–127.77) | <0.001 | 18.14 (2.62–125.71) | 0.003 |
| Anxiety disorder | 3.17 (1.29–7.82) | 0.012 | ||
| MDD | 6.81 (2.50–18.54) | <0.001 | 6.62 (1.72–25.44) | 0.006 |
| Current psychotherapist | 8.87 (2.00–39.38) | 0.004 | ||
| Psychiatrist integrated with primary care | 5.62 (1.06–29.74) | 0.042 | ||
| Psychiatrist elsewhere | 6.67 (2.61–17.03) | <0.001 | 4.52 (1.26–16.22) | 0.021 |
N = 145. Bold indicates statistical significance (p < 0.05).
Table 9.
Significant Outcomes of Logistic Regression Models on Outpatient Behavioral Health Engagement
| Variable | Bivariate models | Multivariable model | ||
|---|---|---|---|---|
| OR (95% CI) | p | OR (95% CI) | p | |
| Demographics | ||||
| AMAB | 0.45 (0.21–0.95) | 0.037 | ||
| Gender identity-related characteristics | ||||
| Any past medically unmonitored hormone use | 0.17 (0.03–0.96) | 0.045 | ||
| Mental health | ||||
| Anxiety disorder | 23.03 (3.04–174.30) | 0.002 | 15.84 (2.00–125.72) | 0.009 |
| MDD | 18.50 (4.24–80.70) | <0.001 | 10.45 (2.28–47.98) | 0.003 |
| History of suicide attempt | 8.30 (1.07–64.23) | 0.043 | ||
| Current case management utilization | 16.43 (2.16–124.92) | 0.007 | 10.73 (1.32–87.53) | 0.027 |
N = 145. Bold indicates statistical significance (p < 0.05).
Discussion
This study expands our understanding of the prevalence of and factors associated with psychiatric diagnoses (including SUDs), psychiatric acuity, and outpatient behavioral health engagement among transgender and nonbinary adults. To our knowledge, this study is the first to comprehensively examine factors associated with these dimensions of mental health in a cohort of transgender and nonbinary adults. Approximately, one-quarter of participants were nonbinary identified; inclusion of this subgroup reflects the reality of clinical practices. Nonbinary identities in this sample did not have a statistically significant association with any outcome assessed in bivariate or multivariable analyses.
Mental illness burden among transgender and nonbinary people
We found a higher prevalence of psychiatric disorders among transgender and nonbinary adults compared with the general population.46 The 28.3% prevalence of any DSM-5 anxiety disorder in this cohort exceeds the 18.1% prevalence of any DSM-IV anxiety disorder in the U.S. adult population, despite the latter including conditions no longer classified as anxiety disorders under DSM-5, such as PTSD and obsessive–compulsive disorder.47 Fewer studies to date have characterized the epidemiology of anxiety disorders in the general population using DSM-5 criteria, and some have found that DSM-5 criteria may lead to increased prevalences of certain disorders across the general population. It will be important to continue comparing prevalences of psychiatric disorders among transgender and nonbinary communities to prevalences in the general population as studies using DSM-5 criteria continue to emerge.48,49
Compared to the 2.1% prevalence of any personality disorder found in this cohort, the prevalence of any personality disorder in the United States is 9.1%.50 Chart audit may yield a different prevalence than a national survey utilizing diagnostic interviewing.50 In addition, our cohort is a clinical population in primary care, which may differ from a general population (e.g., our sample could be in better health due to accessing more preventive care, or more capable of engaging in care due to fewer personality disorder symptoms).
Nearly 14% of study participants had a history of suicide attempt. Although strikingly high compared with suicide attempt prevalence in the general population (4.6%), this estimate is much lower than the lifetime prevalences exceeding 40% found in recent studies.19,20 Many of these studies used surveys, but this study used chart review to identify past suicide attempts. Patients may have underreported suicide attempts to clinicians and our data may thus underestimate attempts. Alternatively, given that all patients in this study were proactively seeking health care, our sample may not have included patients with severe mental illness that impedes them from accessing care and who are more likely to attempt suicide. Regardless of which findings more accurately reflect suicide attempt prevalence, both results are concerning.
Psychiatric diagnoses, SUDs, psychiatric acuity, and outpatient behavioral health engagement
Psychiatric diagnoses
In this cohort, not working was associated with increased odds of any non-SUD psychiatric diagnosis. This finding is consistent with general population data; a recent report found that 80% of people receiving public mental health services were unemployed.51 Employment may provide increased financial security, purpose, and community.51 Although the existing literature is inconsistent with regard to the differential psychiatric epidemiology of AMAB versus AFAB adults,17 in this cohort, being AFAB was associated with higher odds of non-SUD psychiatric diagnoses, consistent with a recent study which found that trans masculine youth carry ∼25% more mood disorder diagnoses than trans feminine youth.52
Substance use disorders
Factors associated with increased odds of a lifetime history of any SUD included older age at hormone therapy initiation and history of suicide attempt. A relationship between age at hormone therapy initiation and SUDs was also present in our previous study, which examined the same sample and which aimed to characterize predictors of gender-affirming care utilization.44 Although that study was otherwise not overlapping in scope or findings, in it we did find that alcohol use disorder, in particular, significantly predicted older age at hormone initiation. Given the common sample investigated in that study and this study, the related findings are not surprising. This study newly characterizes the association as being broader than with alcohol use alone, and it may be explained by decreased at-risk substance use after accessing gender-affirming care.7 Transgender and nonbinary adults with access to gender-affirming hormones at an earlier age may experience less gender dysphoria and develop more adaptive coping skills.7,20,44,53,54 Alternatively, participants with SUDs may have more behavioral disorganization or encounter gatekeeping by clinicians, both of which may impede linkage to gender-affirming medical care.44
The relationship between suicide attempts and increased odds of any SUD is complex. SUDs are second only to mood disorders as a risk factor for suicide.55 In addition, transgender and nonbinary people may experience lifelong minority stress associated with their gender identity, which can impede development of adaptive coping strategies and increase likelihood of attempting suicide or using substances to manage life stressors.20,53 Provision of gender-affirming care that addresses social determinants of health may decrease both SUDs and suicide risk among transgender and nonbinary people.56
Psychiatric acuity
BPAD, a psychiatric disorder thought to be less directly driven by longitudinal environmental influences, was not associated with increased psychiatric acuity in our study.57 This finding was surprising, as many studies have found increased suicidal ideation and attempts among individuals with BPAD.58–63 The absence of integration of psychiatrists into primary care was associated with increased odds of acuity, whereas integration of psychiatrists with primary care was not. Taken together, these findings have vital implications for designing health care systems. Psychiatric service integration into primary care through collaborative care models could help improve mental health for transgender and nonbinary patients, as has been demonstrated in other patient populations.64–67
Outpatient behavioral health engagement
Case management utilization was significantly associated with outpatient behavioral health engagement. This suggests that case management may improve engagement with behavioral health care within this population. Alternatively, this association could reflect the increased likelihood for behavioral health providers to work in teams and engage patients with case management services. Gender minority people disproportionately experience social determinants such as poverty, housing instability, and unemployment that impact health adversely; thus, they may particularly benefit from case management support.53
Outpatient behavioral health engagement was associated with anxiety disorders and MDD. Anxiety and depression are highly prevalent among transgender adults,68 and it is therefore heartening that this population appears ready to engage in behavioral health care when an inclusive, affirming environment is accessible. Although data are not available on gender minority people, sexual minority people are three times more likely to use group therapy compared with heterosexual-identified people, and four times more likely to use psychopharmacology compared with heterosexual-identified people.69 Given the similar mechanisms of action for minority stress on both sexual and gender minority adults, we are hopeful that transgender and nonbinary patients will access needed care if affirming services are offered.
Limitations
Limitations of this study should be addressed. First, the cross-sectional study design does not allow for causal inference regarding relationships between predictor and outcome variables. Second, the study sample is from one LGBTQ-specialized urban health center that provides care to a primarily Caucasian patient population, limiting generalizability. The sample was drawn from an FQHC, which by definition cares for a patient population with lower socioeconomic status that may therefore have a higher burden of mental illness.70 We were limited in our comparison of this cohort with regard to personality disorder diagnoses; we made the comparison to a national sample, as personality disorder prevalence has not yet been studied within health centers. With one investigating psychiatrist carrying out the manual chart audit, we did not determine interrater reliability for diagnostic assessments. In addition, the patient sampling method consisted only of gender minority patients actively engaged in care. The National Transgender Discrimination Survey found that 30.8% of transgender participants delayed or did not seek care due to discrimination.53,71 Thus, this sample may represent a cohort that has experienced less discrimination than the general transgender adult population.
Moreover, incomplete charting, clinical information excluded from patient notes, subjective documentation, and variability in providers' diagnostic biases and documentation practices represent important sources of potential bias that could not be controlled for in this study design.38,39,72–75 Several factors known to be predictive of psychiatric diagnoses, including emotional, physical, and sexual trauma, were not included in our models due to inconsistent medical record documentation for these variables. Future research is needed to replicate these findings and expand upon them with adjustment for other covariates and confounders not modeled in this study.
Although this study emphasized rigorous confirmation of patient-level information by manual EHR audit of clinical encounter notes from a smaller subset of the health center's overall transgender and nonbinary patients, larger samples that allow for disaggregation of current gender identity and sex assigned at birth are needed to assess subgroup differences in psychiatric diagnoses among gender minorities. For example, future studies ought to stratify by current gender identity and sex assigned at birth, and by more granular gender identifications among nonbinary people. Stratification based on current gender identity could help deepen our understanding of how current lived experience and physiology relate to mental health, as distinct from associations with sex assigned at birth that may include the influence of genetic factors, early hormonal exposure, and past lived experience. Similarly, larger samples may enable investigation of more complex dynamics among variables than we were able to study, such as level-of-care escalation extent and frequency, the interactions of SUDs with other factors, and additional variables related to stigma and minority stress. Finally, approximately one-quarter of participants were excluded due to missing data, making the data vulnerable to sample variance.
Conclusion
Our findings point to potential clinical- and systems-level strategies that may mitigate risk factors and promote protective factors that influence psychiatric morbidity among gender minority people. Novel findings include the relationships of systems-level factors to outcomes, such as the association of lack of psychiatric service integration with greater psychiatric acuity, and of case management utilization with greater outpatient behavioral health engagement.
Future work ought to refine tailored, gender-affirming mental health interventions for transgender and nonbinary adults. To expand dissemination and accessibility of such interventions, it will also be critical to continue developing novel models of screening, integration, and collaboration, with staffing by the full continuum of mental health clinicians, including psychiatric mental health nurse practitioners and clinical social workers.76–78 Although our cross-sectional study assessed risk and protective factors at individual and systems levels associated with psychiatric morbidity in a cohort of transgender and nonbinary adults, we did not study community-level factors or mechanisms by which these factors influence mental health. Future research ought also to focus on elucidating mechanisms by which stigma and other stressors contribute to adverse mental health in these marginalized groups.79 Such studies can inform mental health-focused prevention and treatment strategies tailored for gender minority people.
Acknowledgments
This research was supported by grant U30CS22742 from the Health Resources and Services Administration Bureau of Primary Health Care, grant R34MH104072 from the National Institute of Mental Health (Bethesda, MD), and grant K24 DA022288 from the National Institute on Drug Abuse.
Disclaimer
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Author Disclosure Statement
No competing financial interests exist.
References
- 1. UCSF Center of Excellence for Transgender Health: Guidelines for the primary and gender-affirming care of transgender and gender nonbinary people: Terminology and definitions. Available at http://transhealth.ucsf.edu/trans?page=guidelines-terminology Accessed March29, 2017
- 2. Mayer KH, Bradford JB, Makadon HJ, et al. : Sexual and gender minority health: What we know and what needs to be done. Am J Public Health 2008;98:989–995 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Flores AR, Herman JL, Gates GJ, Brown TNT: How Many Adults Identify as Transgender in the United States? Los Angeles, CA: The Williams Institute, 2016 [Google Scholar]
- 4. National LGBT Health Education Center: Glossary of LGBT Terms for Health Care Teams. 2017. Available at https://www.lgbthealtheducation.org/wp-content/uploads/2018/03/Glossary-2018-English-update-1.pdf Accessed November28, 2017
- 5. Reisner SL, Katz-Wise SL, Gordon AR, et al. : Social epidemiology of depression and anxiety by gender identity. J Adolesc Health 2016;59:203–208 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Aparicio-García ME, Díaz-Ramiro EM, Rubio-Valdehita S, et al. : Health and well-being of cisgender, transgender and non-binary young people. Int J Environ Res Public Health 2018;15:E2133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Keuroghlian AS, Reisner SL, White JM, Weiss RD: Substance use and treatment of substance use disorders in a community sample of transgender adults. Drug Alcohol Depend 2015;152:139–146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Flentje A, Bacca CL, Cochran BN: Missing data in substance abuse research? Researchers' reporting practices of sexual orientation and gender identity. Drug Alcohol Depend 2015;147:280–284 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. McDermott E, Hughes E, Rawlings V: Norms and normalisation: Understanding lesbian, gay, bisexual, transgender and queer youth, suicidality and help-seeking. Cult Health Sex 2018;20:156–172 [DOI] [PubMed] [Google Scholar]
- 10. Matsuzaka S: Transgressing gender norms in addiction treatment: Transgender rights to access within gender-segregated facilities. J Ethn Subst Abuse 2018;17:420–433 [DOI] [PubMed] [Google Scholar]
- 11. Freese R, Ott MQ, Rood BA, et al. : Distinct coping profiles are associated with mental health differences in transgender and gender nonconforming adults. J Clin Psychol 2018;74:136–146 [DOI] [PubMed] [Google Scholar]
- 12. Kussin-Shoptaw AL, Fletcher JB, Reback CJ: Physical and/or sexual abuse is associated with increased psychological and emotional distress among transgender women. LGBT Health 2017;4:268–274 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Bockting WO, Miner MH, Swinburne Romine RE, et al. : Stigma, mental health, and resilience in an online sample of the US transgender population. Am J Public Health 2013;103:943–951 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Schulman JK, Erickson-Schroth L: Mental health in sexual minority and transgender women. Psychiatr Clin North Am 2017;40:309–319 [DOI] [PubMed] [Google Scholar]
- 15. Flentje A, Leon A, Carrico A, et al. : Mental and physical health among homeless sexual and gender minorities in a major urban US city. J Urban Health 2016;93:997–1009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Figueiredo AR, Abreu T: Suicide among LGBT individuals. Eur Psychiatry 2015;30:1815 [Google Scholar]
- 17. Wolford-Clevenger C, Frantell K, Smith PN, et al. : Correlates of suicide ideation and behaviors among transgender people: A systematic review guided by ideation-to-action theory. Clin Psychol Rev 2018;63:93–105 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Substance Abuse and Mental Health Services Administration: Mental and substance use disorders. 2014. Available at https://www.samhsa.gov/disorders Accessed August25, 2017
- 19. McNeil J, Ellis SJ, Eccles FJR: Suicide in trans populations: A systematic review of prevalence and correlates. Psychol Sex Orientat Gend Divers 2017;4:341–353 [Google Scholar]
- 20. James SE, Herman JL, Rankin S, et al. : The Report of the 2015 U.S. Transgender Survey Washington, DC: National Center for Transgender Equality, 2016 [Google Scholar]
- 21. Hatzenbuehler ML: How does sexual minority stigma “get under the skin”? A psychological mediation framework. Psychol Bull 2009;135:707–730 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Roberts AL, Rosario M, Corliss HL, et al. : Childhood gender nonconformity: A risk indicator for childhood abuse and posttraumatic stress in youth. Pediatrics 2012;129:410–417 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Reisner SL, White Hughto JM, Gamarel KE, et al. : Discriminatory experiences associated with posttraumatic stress disorder symptoms among transgender adults. J Couns Psychol 2016;63:509–519 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Harrison PJ, Geddes JR, Tunbridge EM: The emerging neurobiology of bipolar disorder. Trends Neurosci 2018;41:18–30 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Reichborn-Kjennerud T: The genetic epidemiology of personality disorders. Dialogues Clin Neurosci 2010;12:103–114 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Dhejne C, Van Vlerken R, Heylens G, Arcelus J: Mental health and gender dysphoria: A review of the literature. Int Rev Psychiatry 2016;28:44–57 [DOI] [PubMed] [Google Scholar]
- 27. Heylens G, Elaut E, Kreukels BP, et al. : Psychiatric characteristics in transsexual individuals: Multicentre study in four European countries. Br J Psychiatry 2014;204:151–156 [DOI] [PubMed] [Google Scholar]
- 28. Flentje A, Heck NC, Sorensen JL: Characteristics of transgender individuals entering substance abuse treatment. Addict Behav 2014;39:969–975 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Willging CE, Salvador M, Kano M: Pragmatic help seeking: How sexual and gender minority groups access mental health care in a rural state. Psychiatr Serv 2006;57:871–874 [DOI] [PubMed] [Google Scholar]
- 30. Klotzbaugh R, Glover E: A lesbian, gay, bisexual and transgender dedicated inpatient psychiatric unit in rural New England: A descriptive analysis in demographics, service utilisation and needs. J Clin Nurs 2016;25:3570–3576 [DOI] [PubMed] [Google Scholar]
- 31. Blosnich JR, Brown GR, Shipherd Phd JC, et al. : Prevalence of gender identity disorder and suicide risk among transgender veterans utilizing veterans health administration care. Am J Public Health 2013;103:e27–e32 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Cochran BN, Cauce AM: Characteristics of lesbian, gay, bisexual, and transgender individuals entering substance abuse treatment. J Subst Abuse Treat 2006;30:135–146 [DOI] [PubMed] [Google Scholar]
- 33. O'Cleirigh C, Dale SK, Elsesser S, et al. : Sexual minority specific and related traumatic experiences are associated with increased risk for smoking among gay and bisexual men. J Psychosom Res 2015;78:472–477 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Ross LE, O'Gorman L, MacLeod MA, et al. : Bisexuality, poverty and mental health: A mixed methods analysis. Soc Sci Med 2016;156:64–72 [DOI] [PubMed] [Google Scholar]
- 35. Mayer K, Appelbaum J, Rogers T, et al. : The evolution of the Fenway Community Health model. Am J Public Health 2001;91:892–894 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Reisner SL, Bradford J, Hopwood R, et al. : Comprehensive transgender healthcare: The gender affirming clinical and public health model of Fenway Health. J Urban Health 2015;92:584–592 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Thompson SK: Simple random sampling. In: Sampling, 3rd ed. Hoboken, NJ: John Wiley& Sons, Inc., 2012, pp 9–37 [Google Scholar]
- 38. Myers L, Stevens J: Using EHR to conduct outcome and health services research. In: Secondary Analysis of Electronic Health Records. Cambridge, MA: Springer International Publishing, 2016, pp 61–70 [PubMed] [Google Scholar]
- 39. Casey JA, Schwartz BS, Stewart WF, Adler NE. Using electronic health records for population health research: A review of methods and applications. Annu Rev Public Health 2016;37:61–81 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 5th ed. (DSM-5). Arlington, VA, 2013 [Google Scholar]
- 41. Ramrakha S, Paul C, Bell ML, et al. : The relationship between multiple sex partners and anxiety, depression, and substance dependence disorders: A cohort study. Arch Sex Behav 2013;42:863–872 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Byers AL, Lai AX, Nelson C, Yaffe K: Predictors of mental health services use across the life course among racially-ethnically diverse adults. Am J Geriatr Psychiatry 2017;25:1213–1222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Kim P, Evans GW, Chen E, et al. : How socioeconomic disadvantages get under the skin and into the brain to influence health development across the lifespan. In: Handbook of Life Course Health Development. Edited by Halfon N, Forrest CB, Lerner RM, Faustman EM. Cham, Switzerland:Springer International Publishing, 2018, pp 463–497 [PubMed] [Google Scholar]
- 44. Beckwith N, Reisner SL, Zaslow S, et al. : Factors associated with gender-affirming surgery and age of hormone therapy initiation among transgender adults. Transgend Health 2017;2:156–164 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Afifi A, Clark VA, May S: Computer-Aided Multivariate Analysis, 4th ed. Boca Raton, FL: Chapman and Hall/CRC, 2003 [Google Scholar]
- 46. National Institute of Mental Health. Mental illness. Available at https://www.nimh.nih.gov/health/statistics/mental-illness.shtml#part_154785 Accessed April6, 2018
- 47. Kessler RC, Berglund P, Demler O, et al. : Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 2005;62:593–602 [DOI] [PubMed] [Google Scholar]
- 48. Karlsson B, Sigström R, Östling S, et al. : DSM-IV and DSM-5 prevalence of social anxiety disorder in a population sample of older people. Am J Geriatr Psychiatry 2016;24:1237–1245 [DOI] [PubMed] [Google Scholar]
- 49. Ruscio AM, Hallion LS, Lim CCW, et al. : Cross-sectional comparison of the epidemiology of DSM-5 generalized anxiety disorder across the globe. JAMA Psychiatry 2017;74:465–475 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Lenzenweger MF, Lane MC, Loranger AW, Kessler RC: DSM-IV personality disorders in the National Comorbidity Survey Replication. Biol Psychiatry 2007;62:553–564 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. National Alliance on Mental Illness (NAMI): Road to Recovery: Employment and Mental Illness. Arlington, VA: National Alliance on Mental Illness (NAMI), 2014 [Google Scholar]
- 52. Becerra-Culqui TA, Liu Y, Nash R, et al. : Mental health of transgender and gender nonconforming youth compared with their peers. Pediatrics 2018:e20173845. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Grant JM, Mottet LA, Tanis J, et al. : Injustice at Every Turn: A Report of the National Transgender Discrimination Survey. Washington, DC: National Center for Transgender Equality and National Gay and Lesbian Task Force, 2011 [Google Scholar]
- 54. 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;1:21–31 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Center for Substance Abuse Treatment: Substance Abuse and Suicide Prevention: Evidence and Implications—A White Paper. DHHS Pub. No. SMA-08-4352. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2008 [Google Scholar]
- 56. Rowe C, Santos GM, McFarland W, Wilson EC: Prevalence and correlates of substance use among trans female youth ages 16–24 years in the San Francisco Bay Area. Drug Alcohol Depend 2015;147:160–166 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Burmeister M, McInnis MG, Zöllner S: Psychiatric genetics: Progress amid controversy. Nat Rev Genet 2008;9:527–540 [DOI] [PubMed] [Google Scholar]
- 58. Dennehy EB, Marangell LB, Allen MH, et al. : Suicide and suicide attempts in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). J Affect Disord 2011;133:423–427 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59. Marangell LB, Bauer MS, Dennehy EB, et al. : Prospective predictors of suicide and suicide attempts in 1,556 patients with bipolar disorders followed for up to 2 years. Bipolar Disord 2006;8(5 Pt 2):566–575 [DOI] [PubMed] [Google Scholar]
- 60. Goetz I, Tohen M, Reed C, et al. : Functional impairment in patients with mania: Baseline results of the EMBLEM study. Bipolar Disord 2007;9:45–52 [DOI] [PubMed] [Google Scholar]
- 61. Hawton K, Sutton L, Haw C, et al. : Suicide and attempted suicide in bipolar disorder: A systematic review of risk factors. J Clin Psychiatry 2005;66:693–704 [DOI] [PubMed] [Google Scholar]
- 62. Angst J, Angst F, Gerber-Werder R, Gamma A: Suicide in 406 mood-disorder patients with and without long-term medication: A 40 to 44 years' follow-up. Arch Suicide Res 2005;9:279–300 [DOI] [PubMed] [Google Scholar]
- 63. Osby U, Brandt L, Correia N, et al. : Excess mortality in bipolar and unipolar disorder in Sweden. Arch Gen Psychiatry 2001;58:844–850 [DOI] [PubMed] [Google Scholar]
- 64. Tully PJ, Baumeister H: Collaborative care for the treatment of comorbid depression and coronary heart disease: A systematic review and meta-analysis protocol. Syst Rev 2014;3:127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65. Unützer J, Katon W, Callahan CM, et al. : Collaborative care management of late-life depression in the primary care setting: A randomized controlled trial. JAMA 2002;288:2836–2845 [DOI] [PubMed] [Google Scholar]
- 66. Dowshen N, Lee S, Franklin J, et al. : Access to medical and mental health services across the HIV care continuum among young transgender women: A qualitative study. Transgend Health 2017;2:81–90 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67. Sevelius JM, Patouhas E, Keatley JG, Johnson MO: Barriers and facilitators to engagement and retention in care among transgender women living with human immunodeficiency virus. Ann Behav Med 2014;47:5–16 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68. Budge SL, Adelson JL, Howard KA: Anxiety and depression in transgender individuals: The roles of transition status, loss, social support, and coping. J Consult Clin Psychol 2013;81:545–557 [DOI] [PubMed] [Google Scholar]
- 69. Cochran SD, Sullivan JG, Mays VM: Prevalence of mental disorders, psychological distress, and mental health services use among lesbian, gay, and bisexual adults in the United States. J Consult Clin Psychol 2003;71:53–61 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70. Hudson CG: Socioeconomic status and mental illness: Tests of the social causation and selection hypotheses. Am J Orthopsychiatry 2005;75:3–18 [DOI] [PubMed] [Google Scholar]
- 71. Jaffee KD, Shires DA, Stroumsa D: Discrimination and delayed health care among transgender women and men: Implications for improving medical education and health care delivery. Med Care 2016;54:1010–1016 [DOI] [PubMed] [Google Scholar]
- 72. Dworkin RJ: Hidden bias in the use of archival data. Eval Health Prof 1987;10:173–185 [DOI] [PubMed] [Google Scholar]
- 73. Gearing RE, Mian IA, Barber J, Ickowicz A: A methodology for conducting retrospective chart review research in child and adolescent psychiatry. J Can Acad Child Adolesc Psychiatry 2006;15:126–134 [PMC free article] [PubMed] [Google Scholar]
- 74. Hess DR: Retrospective studies and chart reviews. Respir Care 2004;49:1171–1174 [PubMed] [Google Scholar]
- 75. Pan L, Fergusson D, Schweitzer I, Hebert PC: Ensuring high accuracy of data abstracted from patient charts: The use of a standardized medical record as a training tool. J Clin Epidemiol 2005;58:918–923 [DOI] [PubMed] [Google Scholar]
- 76. Reisner SL, Radix A, Deutsch MB: Integrated and gender-affirming transgender clinical care and research. J Acquir Immune Defic Syndr 2016;72 Suppl 3:S235–S242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 77. Kaigle A, Sawan-Garcia R, Firek A: Approach to the provision of transgender health care in a veteran population. Ment Health Clin 2018;7:176–180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 78. McDowell A, Bower KM: Transgender health care for nurses: An innovative approach to diversifying nursing curricula to address health inequities. J Nurs Educ 2016;55:476–479 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 79. Reisner SL, Pardo ST, Gamarel KE, et al. : Substance use to cope with stigma in healthcare among U.S. female-to-male trans masculine adults. LGBT Health 2015;2:324–332 [DOI] [PMC free article] [PubMed] [Google Scholar]