Decreased Anogenital Gonorrhea and Chlamydia in Transgender and Gender Diverse Primary Care Patients Receiving Gender-Affirming Hormone Therapy

Abstract

Background

Transgender and gender diverse (TGD) adults in the U.S. experience health disparities, including in anogenital sexually transmitted infections (STI). Gender-affirming hormone therapy (GAHT) is known to be medically necessary and improve health. Few studies have assessed the effect of GAHT on STI diagnoses.

Objective

To evaluate the effect of GAHT delivered in primary care as an intervention to improve STI outcomes for TGD adults.

Design

LEGACY is a longitudinal, multisite cohort study of adult TGD primary care patients from two federally qualified community health centers in Boston, MA, and New York, NY.

Participants

Electronic health record data for eligible adult TGD patients contributed to the LEGACY research data warehouse (RDW). A total of 6330 LEGACY RDW patients were followed from 2016 to 2019, with 2555 patients providing STI testing data.

Main Measures

GAHT exposure was being prescribed hormones, and the clinical outcome was anogenital gonorrhea or chlamydia diagnoses. Log-Poisson generalized estimating equations assessed the effect of prescription GAHT on primary outcomes, adjusting for age, race, ethnicity, gender identity, poverty level, health insurance, clinical site, and cohort years.

Key Results

The median age was 28 years (IQR = 13); the racial breakdown was 20.4% Black, 8.1% Multiracial, 6.9% Asian/Pacific Islander, 1.8% Other; 62.8% White; 21.3% Hispanic/Latinx; 47.0% were assigned female at birth, and 16.0% identified as nonbinary. 86.3% were prescribed hormones. Among those tested, the percentage of patients with a positive anogenital STI diagnosis ranged annually from 10.0 to 12.5% between 2016 and 2019. GAHT prescription was associated with a significant reduction in the risk of anogenital STI diagnosis (aRR = 0.75; 95% CI = 0.59–0.96) over follow-up.

Conclusions

GAHT delivered in primary care was associated with less STI morbidity in this TGD cohort over follow-up. Patients may benefit from individualized and tailored clinical care alongside GAHT to optimize STI outcomes.

INTRODUCTION

In the United States (U.S.), transgender and gender diverse (TGD) adults—those whose gender identity does not correspond to their sex assigned at birth—are highly burdened by sexually transmitted infections (STIs).¹ An international review of STI studies in TGD people found gonorrhea prevalence estimates ranging from 2.1 to 19.1% in transgender women and 0 to 10.5% in transgender men, and rates of chlamydia from 2.7 to 24.7% in transgender women and 1.2 to 11.1% in transgender men.¹ Routine STI surveillance systems in the U.S. capture limited demographic information, often restricted to age, sex, race, and ethnicity, leaving gaps in reporting of STIs for TGD people.² A 2019 study of 626 TGD people accessing publicly funded STI services in 6 U.S. cities found a laboratory-confirmed gonorrhea prevalence of 13.1% in transgender women and 10.5% in transgender men, and chlamydia prevalence of 13.1% and 7.7%, respectively.³ Black and Latinx/e TGD people are particularly affected by STIs.⁴ Data are lacking about STIs in nonbinary TGD people.² In addition to the morbidity that STIs cause, they are a biological marker of HIV sexual risk and increase HIV acquisition or transmission vulnerability.⁵ In the context of the population’s high HIV epidemic burden,⁶ TGD people are a priority population for STI prevention and care. Multiple individual (e.g., demographic), interpersonal (e.g., HIV serostatus communication with partners), and structural (e.g., stigma, racism) factors increase STI risks in TGD people.⁷ These vulnerabilities are driven by and associated with barriers that limit access to gender-affirming STI prevention, diagnosis, and treatment services. Barriers may include mistrust of providers and medical settings due to prior negative healthcare experiences, or prioritization of other heath needs over STI prevention, such as receipt of medical gender affirmation services.⁸

Gender-affirming hormone therapy (GAHT) is a medically necessary treatment shown to improve psychological functioning and quality of life for TGD adults.¹ GAHT is a critical health concern for many TGD patients.^{9, 10} It is not yet known whether GAHT decreases STI incidence over time in TGD adult patients. This is because prior studies providing the best evidence of GAHT’s clinical effectiveness have not examined STI outcomes.^{11, 12} Lack of requisite knowledge concerning GAHT and STI outcomes impedes the design, implementation, evaluation, and funding of healthcare and service delivery models that may reduce STI disparities for TGD people. Integrating GAHT with STI prevention, diagnosis, and treatment may improve clinical outcomes for TGD people.¹³ The current project sought to fill these evidence gaps.

The objective of this study was to evaluate the effect of GAHT delivered in primary care as an intervention to improve STI outcomes for TGD adults. We assessed whether GAHT was associated with reduced rates of past 12-month anogenital gonococcal or chlamydia positivity for TGD patients over 48 months of follow-up in a cohort of adult TGD primary care patients diverse in terms of age, race, ethnicity, gender identity, and HIV serostatus.

METHODS

Participants and Procedures

The LEGACY Project is a prospective, multisite observational cohort of adult TGD primary care patients from two federally qualified health centers (FQHCs) in Boston, MA, and New York City, NY. Both FQHCs have long histories of providing culturally responsive and affirming healthcare for TGD people, and both use an informed consent model to provide GAHT.¹⁴ LEGACY cohort eligibility criteria were (1) being age ≥ 18 years (verified via electronic health record [EHR]), (2) having a gender identity different from sex assigned at birth (verified via a two-step method cross-categorizing sex and gender identity reported on patient registration and/or ICD-10 code of F64.0–9),^{15, 16} (3) being a primary care patient defined as having a medical visit within the past 12 months, and (4) having a signed patient consent form on file acknowledging that patient data may be utilized for research and no research exclusion documented in the patient chart. EHR data for eligible adult TGD patients contributed to the LEGACY research data warehouse (RDW), a HIPAA-limited dataset. De-identified EHR data (e.g., provider-documented diagnoses, laboratory data) were extracted from the EHR every 6 months. Patients enrolled in the LEGACY RDW were not individually compensated. A detailed protocol has been published elsewhere.¹⁷

A total of N = 6330 LEGACY RDW patients were followed from 2016 to 2019. During that period, GAHT services were delivered as part of integrated primary care according to the same established standards and guidelines by both Fenway Health and Callen-Lorde.^18–20 We compared 12-month STI outcomes in groups who did and did not receive GAHT across 48 months of follow-up. All study procedures were approved by the Fenway Health Institutional Review Board (FWA00000145), which provided single IRB review for this study. The IRB granted a waiver of written consent to allow automatic enrollment of all existing and new TGD adult patients into the LEGACY RDW who met study eligibility criteria.

Patient Engagement, Community Advisory Board, and Scientific Advisory Board

This project was conceptualized by and for TGD patients, with patients and stakeholders as key personnel. A “participatory population perspective” was used to actively engage communities, ensure patient-centeredness, and integrate the needs and concerns of patients and key stakeholders.²¹ Formative qualitative research was conducted with patients to inform cohort procedures.²² A Community Advisory Board (CAB; 7 TGD individuals) and Scientific Advisory Board (SAB; 7 researchers and/or medical providers) were actively engaged with the research team and provided study input. Each board met at least twice per year and members were compensated $50 per meeting.

Measures

Clinical Outcome: Anogenital STI Diagnosis of Gonorrhea and/or Chlamydia, Last 12 Months

The clinical outcome of interest was anogenital STI diagnosis of gonorrhea and/or chlamydia, last 12 months. The outcome was coded as any positive anogenital STI test (rectal, urethral, cervicovaginal) vs. none based on the most recent STI test result in the last 12 months (any positive test within the year based on the last result date). Biological specimens were collected for bacterial STI testing from relevant anatomical sites of participants who had clinical indications, as determined by their medical provider. Multisite samples were taken based on sexual activity and patient anatomy; however, if a patient tested positive for an infection at more than one site, this was counted as a single infection for analyses. Urine, cervicovaginal, and anorectal swabs were provider- or self-collected (depending on participant preference) to test for Neisseria gonorrhoeae and Chlamydia trachomatis via the APTIMA™ COMBO 2 Assay (Gen-Probe; > 95.2% sensitivity and > 96.8% specificity). Test results were extracted from patient charts.

Exposure: Gender-Affirming Hormone Therapy (GAHT)

The exposure variable of GAHT was coded as the binary variable of prescribed GAHT (yes, no) in EHR data. The variable was operationalized using the date of the first GAHT prescription and the date of the last GAHT prescription in each year of observation. We dichotomously coded GAHT within the year vs. no GAHT during the year. Hormone prescription types included in this variable were pubertal blockers (e.g., leuprolide), antiandrogen (e.g., spironolactone), estrogen (e.g., estradiol), progesterone (e.g., micronized progesterone), and testosterone (e.g., testosterone gel or injection).

Covariates and Confounders

Baseline age was derived from age within the year of the observation and coded into seven categories: ages 18–24, 25–29, 30–34, 35–39, 40–44, 45–49, or 50 + years. Assigned sex at birth and current gender identity were assessed at patient registration and coded as transgender women/female, transgender man/male, nonbinary assigned male sex at birth (AMAB), nonbinary assigned female sex at birth (AFAB), and another gender identity.

Race in the EHR was American Indian/Alaska Native, Another racial identity, Asian, Black/African American, Multiracial, Native Hawaiian/Other Pacific Islander, or White. Hispanic/Latinx ethnicity was assessed as a binary variable (Hispanic/Latinx yes, no). Health insurance was coded as private, public, or uninsured. Federal poverty level (FPL) was operationalized according to U.S. national benchmarks as 0–99%, 100–199%, 200–299%, or 300% + of the poverty level.²³ Cohort years were assessed as the years of the cohort (2016–2019) and ranged from 1 to 4. Site was a binary variable for Fenway Health or Callen-Lorde.

To capture HIV transmission risk, a categorical variable was constructed based on laboratory-confirmed HIV serostatus and prevention-effective adherence. For HIV-uninfected patients: PrEP uptake, last 12 months. This was operationalized as having an active PrEP prescription (Truvada) within the last 12 months vs. none. For HIV-infected patients: virological suppression, last 12 months, was captured from laboratory-confirmed viral load suppression (< 200 copies/mL) vs. not virally suppressed (≥ 200 copies) based on the most recent viral load test in the last 12 months. Laboratory-confirmed HIV serostatus, PrEP prescription, and viral suppression variables were combined into a four-category composite HIV acquisition or transmission risk indicator: HIV-negative-non-PrEP, HIV-negative-PrEP, HIV-positive-not virally suppressed, HIV-positive-virally suppressed. This was operationalized to identify patient subgroups for future interventions in clinical care.

Data Analysis

Analyses were implemented in RStudio (version 2022.07.1) and R (version 4.2.0). Descriptive statistics (frequencies, median, interquartile range [IQR]) were produced for the baseline cohort (N = 6330). An analytic dataset comprised of patients with any STI testing data between 2016 and 2019 (N = 2555; observations = 8598) was used for multiple imputation and statistical modeling. Imputed datasets were generated using the MICE package (Multivariate Imputation via Chained Equations) in R.²⁴ Imputed variables were age, gender identity, race, Hispanic/Latinx ethnicity, insurance, federal poverty level, hormone prescription, PrEP status, HIV status, HIV viral suppression status (i.e., < 200 viral copies), and anogenital STI status (i.e., positive/negative test within the year). Baseline (i.e., 2016) values of these variables were incorporated into the predictor matrix of the imputation model, thus allowing us to establish appropriate temporality in our imputation model (e.g., STI status in 2017 was not used to estimate STI status in 2016).

For longitudinal analyses, generalized estimating equations (GEEs) with log-Poisson link functions²⁵ were used to estimate the marginal or population-average effect of concurrent GAHT on STI diagnosis using robust standard errors.²⁵ Covariate-adjusted GEEs were fit across each of the imputed datasets, and their results were pooled using Rubin’s rules.²⁶ GEE models adjusted for autocorrelation between repeated measures. The QIC statistic was utilized to select the final exchangeable working correlation for models.²⁷ Site and cohort years were conceptualized as design covariates and entered into models as fixed effects. Model 1 included prescribed GAHT only; model 2 added all other variables. Adjusted risk ratios (aRRs) and 95% confidence intervals (95% CIs) were estimated. Finally, a sensitivity analysis was conducted to assess how robust the treatment-outcome association was to potential unmeasured or uncontrolled confounding. We calculated the E-value necessary to explain the observed association between prescribed GAHT and risk of anogenital STI.^{28, 29}

RESULTS

Sample Characteristics

Table 1 presents descriptive characteristics of the baseline cohort in 2016 (N = 6330). The median patient age was 28.0 (IQR = 13.0); approximately half were ages 18–29 years (28.6% ages 18–24, 29.3% ages 25–30). Patients were 36.7% transgender women, 28.9% transgender men, 9.4% nonbinary AFAB, 4.6% nonbinary AMAB, and 20.4% were another unknown transgender identity. Nearly one-quarter (24%) were Black, Indigenous, and Other People of Color, and more than 1 in 5 (21.3%) were Hispanic/Latinx. More than half (54.8%) were at the lowest federal poverty level; 47.5% had private insurance, 36.1% public, and 3.7% had no insurance. The cohort was 38.8% Fenway Health and 61.2% Callen-Lorde patients. The majority (86.3%) were on GAHT at baseline (58.4% estrogen, 51.1% antiandrogens, 43.6% testosterone, 5.0% progesterone, and 0.05% pubertal blockers). GAHT prevalence across years of follow-up is shown in Table 2.

Table 1.

The LEGACY Cohort: Descriptive Characteristics of Transgender and Gender Diverse Patients at Baseline in 2016 (N = 6330)

Characteristic
Clinical outcome: positive anogenital STI TEST, n (%) +
Yes	256 (4.0)
No	6074 (96.0)
Exposure: prescribed gender-affirming hormone therapy, n (%)^
Yes	5461 (86.3)
No	869 (13.7)
Site, n (%)
Callen-Lorde	3875 (61.2)
Fenway Health	2455 (38.8)
Age
Age in years, median (IQR)	28.0 (13.0)
Age group, n (%)
18–24 years	1813 (28.6)
25–30 years	1853 (29.3)
31–40 years	1446 (22.8)
41–50 years	646 (10.2)
51 + years	572 (9.0)
Gender identity, n (%)
Transgender women	2324 (36.7)
Transgender men	1829 (28.9)
Nonbinary AFAB	592 (9.4)
Nonbinary AMAB	294 (4.6)
Transgender, not specified	1291 (20.4)
Race, n (%)
American Indian or Alaskan Native	44 (0.7)
Another race	50 (0.8)
Asian	305 (4.8)
Black or African American	999 (15.8)
Multiracial	396 (6.3)
Native Hawaiian or Other Pacific Islander	74 (1.2)
White	3476 (54.9)
Unknown	986 (15.6)
Ethnicity, n (%)
Hispanic/Latinx	1347 (21.3)
Non-Hispanic/Latinx	3906 (61.7)
Unknown	1077 (17.0)
Health insurance, n (%)
Private insurance	3008 (47.5)
Public insurance	2288 (36.1)
Uninsured	232 (3.7)
Unknown	802 (12.7)
Federal poverty level, n (%)
0–138%	3470 (54.8)
139–250%	807 (12.7)
≥ 251%	1569 (24.8)
Unknown	484 (7.6)
HIV transmission risk category, n (%)
HIV − and not on PrEP	5811 (91.8)
HIV − and on PrEP	82 (1.3)
HIV + and virally suppressed (< 200 copies)	301 (4.8)
HIV + and non-virally suppressed (≥ 200 copies)	85 (1.3)
Unknown	51 (0.8)

Abbreviations: AMAB, assigned male at birth; AFAB, assigned female at birth; STI, sexually transmitted infection; PrEP, pre-exposure prophylaxis

 + The clinical outcome was any anogenital STI diagnosis of gonorrhea and/or chlamydia in the last 12 months. Of positive anogenital STI tests at baseline (n = 256), 87.5% (n = 224) were rectal, 14.8% (n = 38) urethral, and 4.7% (n = 12) cervicovaginal

^Of patients prescribed gender-affirming hormone therapy at baseline, 58.4% (n = 3189) were prescribed estrogen, 51.1% (n = 2792) antiandrogens, 43.6% (n = 2380) testosterone, 5.0% (n = 275) progesterone, and 0.05% (n = 3) pubertal blockers

Table 2.

The LEGACY Cohort: Gender-Affirming Hormone Therapy Prescription Among Transgender and Gender Diverse Patients by Year, 2016–2019

	Total N in cohort	Prescribed GAHT hormones		Not prescribed GAHT hormones
Cohort years		N	% retained	N	% retained
2016	6330	5461	100.0	869	100.0
2017	5267	4910	89.9	357	41.1
2018	4591	4318	79.1	273	31.4
2019	4181	3949	72.3	232	26.7

Abbreviations: GAHT, gender-affirming hormone therapy

Modeling STI Diagnosis

Among TGD patients who underwent testing for anogenital gonorrhea or chlamydia (N = 2555; 40.4% of cohort patients), between 10.0 and 12.5% had a positive anogenital STI test within the last 12 months (see Table 3 for annual incidences across years 2016–2019). Table 4 displays STI tests and diagnoses (number of STI cases per 1000 tests) overall for anogenital gonorrhea or chlamydia, and stratified by prescribed GAHT, site, age group, gender identity, race, and ethnicity.

Table 3.

The LEGACY Cohort: Annual Incidence of STI Tests and Diagnoses in Transgender and Gender Diverse Patients, 2016–2019

	Cohort N	Tests N	% tested	STI positive		STI negative
Cohort years				N	%	N	%
2016	6330	2555	40.4	256	10.0	2299	90.0
2017	5267	2163	41.1	237	11.0	1926	89.0
2018	4591	2047	44.6	221	10.8	1826	89.2
2019	4181	1833	43.8	229	12.5	1604	87.5

The percentage of patients with a positive/negative anogenital STI test was calculated using the annual number of tests as a denominator. Abbreviations: STI, sexually transmitted infections, anogenital gonococcal or chlamydia positivity

Table 4.

The LEGACY Cohort: Anogenital STI Incidence Among Transgender and Gender Diverse Patients During the Observation Period, 2016–2019

	Total tests	STI positive	STI negative	Cases per 1000 tests
Full cohort (2016–2019)	8598	943	7655	109.7
Prescribed gender-affirming hormone therapy
Yes	7899	857	7042	108.5
No	699	86	613	123.0
Site
Callen-Lorde	6465	807	5658	124.8
Fenway	2133	136	1997	63.8
Age group in years
18–24	1928	234	1694	121.4
25–30	2824	347	2477	122.9
31–40	2347	249	2098	106.1
41–50	963	89	874	92.4
51 +	536	24	512	44.8
Gender identity
Transgender woman	3932	568	3364	144.5
Transgender man	2224	168	2056	75.5
Nonbinary AFAB	739	25	714	33.8
Nonbinary AMAB	334	37	297	110.8
Transgender, not specified	1369	145	1224	105.9
Race
American Indian or Alaskan Native	71	6	65	84.5
Another race	44	2	42	45.5
Asian	404	41	363	101.5
Black or African American	2008	332	1676	165.3
Multiracial	622	86	536	138.3
Native Hawaiian or Other Pacific Islander	157	22	135	140.1
White	3538	222	3316	62.7
Unknown	1754	232	1522	132.3
Ethnicity
Hispanic/Latinx	2611	352	2259	134.8
Non- Hispanic/Latinx	4554	430	4124	94.4
Unknown	1433	161	1272	112.4

Abbreviations: AMAB, assigned male at birth; AFAB, assigned female at birth; STI, sexually transmitted infection, anogenital gonococcal or chlamydia positivity

Modeling analyses assessed the longitudinal effect of GAHT prescription on anogenital STI diagnosis and are displayed in Table 5. Modeling the number of longitudinal observations (N = 8598), TGD patients prescribed GAHT had a statistically significantly lower risk of anogenital gonorrhea or chlamydia diagnosis over follow-up compared to those not prescribed GAHT (aRR = 0.75; 95% CI = 0.59–0.96). Lower rates of incident anogenital gonococcal or chlamydia positivity were also seen in older age groups (25–30, 31–40, 41–50, and 51 +) relative to the youngest age group, ages 18–24 (all p < 0.05), and for transgender men (aRR = 0.72, CI = 0.59–0.87) and nonbinary AFAB (aRR = 0.42, CI = 0.26–0.67) compared with transgender women.

Table 5.

The LEGACY Cohort: Results from Longitudinal Multivariable Log-Poisson Models for Positive Anogenital STI Diagnosis Among Transgender and Gender Diverse Patients, 2016–2019 (Patients N = 2555; Number of Observations N = 8598)

	Model 1 +		Model 2
	aRR	95% CI	aRR	95% CI
Prescribed gender-affirming hormone therapy
Yes	0.88	(0.68, 1.12)	0.75	(0.59, 0.96)
No	Ref		Ref
Site
Callen-Lorde	1.65	(1.34, 2.04)	1.30	(1.04, 1.64)
Fenway	Ref		Ref
Cohort years, continuous	1.08	(1.03, 1.14)	1.07	(1.02, 1.13)
Age group in years, n (%)
18–24 years			Ref
25–30 years			0.90	(0.77, 1.06)
31–40 years			0.68	(0.55, 0.83)
41–50 years			0.52	(0.39, 0.70)
51 + years			0.39	(0.24, 0.63)
Gender identity, n (%)
Transgender woman			Ref
Transgender man			0.72	(0.59, 0.87)
Nonbinary AFAB			0.42	(0.26, 0.67)
Nonbinary AMAB			0.92	(0.62, 1.35)
Race, n (%)
American Indian or Alaskan Native			1.13	(0.60, 2.16)
Another race			0.88	(0.26, 2.93)
Asian			1.13	(0.78, 1.65)
Black or African American			1.53	(1.25, 1.86)
Multiracial			1.56	(1.19, 2.04)
Native Hawaiian or Other Pacific Islander			1.23	(0.78, 1.95)
White			Ref
Ethnicity, n (%)
Hispanic/Latinx			1.07	(0.90, 1.27)
Non-Hispanic/Latinx			Ref
Health insurance, n (%)
Private insurance			Ref
Public insurance			1.11	(0.91, 1.34)
Uninsured			1.07	(0.74, 1.53)
Federal poverty level, n (%)
0–138%			1.04	(0.81, 1.33)
139–250%			1.06	(0.79, 1.41)
≥ 251%			Ref
HIV transmission risk category, n (%)
HIV − and not on PrEP			Ref
HIV − and on PrEP			1.89	(1.42, 2.50)
HIV + and virally suppressed (< 200 copies)			1.67	(1.37, 2.03)
HIV + and non-virally suppressed (≥ 200 copies)			2.09	(1.61, 2.72)

Analyses were restricted to patients with anogenital STI testing data (patients N = 2555; number of observations N = 8598). Abbreviations: aRR, adjusted risk ratio; 95% CI, 95% confidence interval; Ref, reference; AMAB, assigned male at birth; AFAB, assigned female at birth; STI, sexually transmitted infection, anogenital gonococcal or chlamydia positivity

Bolded values indicate statistical significance at α = 0.05

 + Generalized estimating equation (GEE) procedures were used to model the outcome of positive anogenital STI test (yes, no) and adjust for autocorrelation between repeated measures. Site and cohort years (number of years contributed to the cohort, continuous) were conceived of as design covariates and entered as fixed effects in models. Model 1 included prescribed gender-affirming hormone therapy. Model 2 included prescribed gender-affirming hormone therapy, age group in years, gender identity, race, ethnicity, health insurance, federal poverty level, and HIV transmission risk category

Higher rates of anogenital gonorrhea or chlamydia diagnosis were found in Black or African American (aRR = 1.53, CI = 1.25–1.86) and Multiracial (aRR = 1.56, CI = 1.19–2.04) compared to White patients; and HIV-negative-PrEP prescribed (aRR = 1.89, CI = 1.42–2.50), HIV-positive virally suppressed (aRR = 1.67, CI = 1.37–2.03), and HIV-positive non-virally suppressed (aRR = 2.09, CI = 1.61–2.72) each compared to HIV-negative-non-PrEP prescribed patients. Callen-Lorde patients had higher rates of anogenital gonococcal or chlamydia positivity compared to Fenway Health patients.

E-value Sensitivity Analysis

In our sensitivity analysis, we calculated an E-value of 2.00 for the covariate-adjusted association between patients’ GAHT and their risk for anogenital gonorrhea or chlamydia (aRR = 0.75; 95% CI = 0.59–0.96). This implies that an unmeasured or uncontrolled confounder, or set of confounders, would need to be associated with a two-fold increase in patients’ risk for an anogenital STI for confounding to completely explain the observed association.

DISCUSSION

This study investigated GAHT and anogenital gonococcal or chlamydia positivity in a diverse multisite cohort of TGD adult primary care patients. GAHT prescribed using informed consent protocols reduced STI morbidity for TGD patients over time in the cohort. TGD patients prescribed GAHT experienced a 25% reduction in the risk of past 12-month anogenital STI diagnosis (a proxy for HIV sexual risk) across follow-up, compared to those not prescribed GAHT. Findings suggest that the provision of GAHT in primary care settings using an informed consent model can improve STI outcomes for TGD patients, though the mechanisms through which GAHT prescription affects STI outcomes are not clear. GAHT is prospectively associated with reductions in mental health–related distress for TGD patients,¹² which in turn may improve health behaviors such as condom use. Receiving care for GAHT may also increase trust with providers and reduce barriers to discussing sexual health concerns, including STI prevention. Future research is needed to understand the pathways through which GAHT affects STI acquisition.

We observed statistically significantly higher anogenital gonorrhea and chlamydia rates in younger age groups, transgender women patients compared to transgender men and nonbinary AFAB patients, and Black and Multiracial relative to White patients. The differential distribution of anogenital gonococcal or chlamydia positivity in the cohort by age and race mirrors national statistics for anogenital STI infection burden.^{4, 30} However, our consideration of gender identity and GAHT in the context of STI acquisition is novel, given many STI health surveillance efforts do not routinely collect this TGD-specific information.² Other prevention-relevant patient characteristics were associated with anogenital gonorrhea or chlamydia diagnosis. HIV-negative-PrEP prescribed patients and HIV-positive virally suppressed patients, and HIV-positive non-virally suppressed patients each had an increased risk of anogenital STI diagnosis compared to HIV-negative-non-PrEP patients. In the context of the clinical outcome of anogenital gonorrhea or chlamydia diagnosis modeled in this study, these are groups vulnerable to HIV acquisition or transmission in need of tailored HIV prevention and care interventions to curb the HIV epidemic.³¹ Findings suggest the continued need for risk reduction interventions in clinical care settings to address STI inequities. Combination interventions that pair GAHT with other services, such as PrEP for HIV prevention (only 1.4% of HIV-negative patients were on PrEP at baseline) and doxycycline post-exposure prophylaxis (doxy-PEP) for STI prevention,³² may hold promise in FQHCs.

Several limitations are important to consider. This cohort was drawn from FQHCs that specialize in providing GAHT through an informed consent model, which allows patients to consent to GAHT as they would any other medication or treatment, thereby minimizing gatekeeping to reduce access barriers. In this study, the effect of hormones on anogenital gonococcal or chlamydia positivity may be confounded by the gender-affirming context in which GAHT prescribing occurred, such as gender-affirming spaces, trained providers, correct name and pronoun use, correct name and gender on health insurance, and linkages to other services such as mental health, gender-affirming surgery, and other subspecialties. Other limitations pertain to weaknesses inherent in a clinical cohort that recruits patients, including clinic patient bias and limited generalizability. This cohort is relatively young and drawn from urban settings, which may limit generalizability. Further, patients may have been healthier than TGD adults not receiving care at specialized FQHCs, which may impact the representative of findings to other TGD patient populations and settings. Other limitations are the self-selection of participants into treatments (e.g., patients self-select GAHT) and the likely underestimation of gonorrhea and chlamydia infections given that STI testing was only done if clinically indicated. Lastly, this study did not assess duration of hormone use; future research is needed to explore the association of duration of GAHT alongside incident STI diagnoses.

Although this study controlled for known, measured confounding in the analysis of GAHT and STIs, there were many factors we were not able to adjust for in this study, leading to the possibility of uncontrolled confounding. For example, we were unable to adjust for the expectations of patients regarding GAHT and whether these expectations were met in prescribing GAHT. In our sensitivity analysis, we calculated an E-value of 2.00, which means that an unmeasured or uncontrolled confounder, or set of confounders, would need to be associated with a two-fold increase in patients’ anogenital STI risk for confounding to completely explain the observed association. In our multivariable model, only being HIV-positive and non-virally suppressed, relative to HIV-negative and not on PrEP, was associated with an adjusted relative risk greater than 2.00 for an anogenital gonorrhea or chlamydia diagnosis. Thus, our results suggest that the covariate-adjusted association between patients’ GAHT and their anogenital STI risk is unlikely to be entirely explained by unobserved confounding.

The LEGACY cohort overcomes many weaknesses of other TGD cohorts, namely, lack of racial and ethnic diversity, restriction to patients in gender clinics, and a high number of nonbinary-identified patients. TGD patients are a population historically excluded from clinical care and research, underscoring the importance of this study and knowledge gained. The longitudinal design allowed for the examination of exposure-outcome associations across 48 months of follow-up, and the EHR dataset was substantially larger than many TGD clinical samples and was diverse in terms of age, gender identity, race, ethnicity, and HIV serostatus. Both FQHCs are real-world implementation settings to ensure the clinical relevance and applicability of research findings. This was patient-centered research conducted with TGD patients and key stakeholders, foregrounding health issues important to them.

CONCLUSIONS

Being prescribed GAHT was longitudinally associated with reduced rates of anogenital gonorrhea or chlamydia diagnosis (a proxy for HIV risk) in this FQHC-based study of diverse TGD patients. Findings lend support to the health-promoting role of GAHT and contribute to the evidence base that GAHT is a medically necessary treatment shown to improve health outcomes longitudinally for TGD people. Results highlight the importance of gender-affirming models of care and access to GAHT for TGD patients as well as opportunities for improving the coordination of primary care with STI services.

Funding

The research reported in this publication was supported by the Patient-Centered Research Outcomes Institute (PCORI) under Award Number AD-2017C1-6569 (PI: Dr. Sari L. Reisner). The content is solely the responsibility of the authors and does not necessarily represent the official views of PCORI. PCORI was not involved in the collection, analysis, or interpretation of study data. 

Declarations

Conflict of Interest:

Authors SLR, JP, and ASK receive royalties from McGraw Hill for editing the textbook, “Transgender and Gender Diverse Health Care: The Fenway Guide.”