Approach to the Patient: Hormonal Therapy in Transgender Adults With Complex Medical Histories

Abstract

While endocrinologists continue to initiate gender-affirming hormone therapy (GAHT) in healthy transgender and gender diverse (TGD) patients, they may also encounter more TGD patients in their clinics with complex medical histories that influence the patient-provider shared decision-making process for initiating or continuing GAHT. The purpose of this Approach to the Patient article is to describe management considerations in 2 adults with thromboembolic disease and 2 adults with low bone mineral density in the setting of feminizing and masculinizing GAHT.

Transgender and gender diverse (TGD) people often initiate gender-affirming hormone therapy (GAHT, see Table 1) (3, 7, 8). Published guidelines provide clinicians with the rationale, evidence, expected outcomes, and management strategies to prescribe safe and effective GAHT (3, 7, 8). However, it is important to have patient-provider discussions summarizing the available data on long-term outcomes related to GAHT and routes of administration. For example, feminizing GAHT (ie, estrogen plus antiandrogen) has been associated with increased risk of venous thromboembolism (VTE), ischemic stroke, and myocardial infarction compared with the general population, depending on estrogen formulation, cohort, and comparison group (9, 10). Masculinizing GAHT (ie, testosterone) does not appear to have similar risk of cardiovascular outcomes as feminizing GAHT, but a recent study showed an increased risk for myocardial infarction among transgender men as compared with cisgender women (10). Regarding bone health, low bone mineral density (BMD) has been seen in TGD patients prior to GAHT initiation and after patients stop GAHT post-gonadectomy, but long-term feminizing and masculinizing GAHT are associated with at least maintenance of BMD (11, 12).

Table 1.

Overview of GAHT regimens and considerations^a

Patient	Category	Name	Recommended dosage ranges	Route	Considerations
Transgender women/transfeminine	Estrogen	17-beta estradiol	2-6 mg once daily (or divided twice daily)	Oral or sublingual	Sublingual 17-beta estradiol has recently been shown to have a higher serum estradiol peak and area under the curve compared to oral administration, but future studies are needed to assess safety and efficacy (1).
		Estradiol valerate	2-6 mg once daily (or divided twice daily)	Oral or sublingual	Not available by itself in the United States.
		Estradiol valerate	5-30 mg once every 2 weeks or 2-10 mg once weeklyb	Intramuscular or subcutaneous	Recent data suggest lower mg doses (3.75 to 4 mg) of both intramuscular and subcutaneous injections are sufficient to achieve guideline-recommended serum estradiol levels (eg, 100-200 pg/mL [367.10-734.20 pmol/L]) (2). Not available in all European countries.
		Estradiol cypionate	5-30 mg once every 2 weeks or 2-10 mg once weeklyb	Intramuscular or subcutaneous
		Estradiol patch	0.025-0.2 mg/day weekly or twice weekly	Transdermal	Use when initiating in adults aged >45 years or elevated risk of/previous thrombosis (3).
		Estradiol gel	Varies based on formulation and strengthc	Transdermal	Consider if adverse effects to the patches and higher risk for thrombosis.
	Antiandrogen (androgen-lowering)	Spironolactone	100-300 mg daily	Oral	Recent studies revealed the prevalence of hyperkalemia in TGD patients taking spironolactone to be low, around 2%, suggesting routine and frequent potassium monitoring may not be necessary except in people aged >45 years or with comorbidities associated with hyperkalemia risk (4, 5).
		Cyproterone acetate	10 mg daily	Oral	Recently, lower doses have been suggested for cyproterone (as opposed to 25, 50, and 100 mg daily), with 10 mg daily equally effective at lowering testosterone and having fewer adverse effects compared to higher doses (6). Not available in the United States.
		Gonadotropin-releasing hormone agonistb	3.75-7.5 mg once monthly (eg, leuprolide or triptorelin) or 11.25 mg once per 3 months (eg, leuprolide or triptorelin) or 22.5 mg once per 6 months (eg, leuprolide triptorelin) or 3.6 mg once per 4 weeks (eg, goserelin)	Intramuscular or subcutaneous	Usually administered in the hospital/clinic setting. May be cost prohibitive for some patients. In Europe, administration can be performed at home with the help of a nurse or general practitioner.
Transgender men/transmasculine	Androgen	Testosterone undecanoate	1000 mg every 12 weeks or 750 mg every 10 weeks	Intramuscular	Administered in the hospital/clinic setting. In Europe, administration can be performed at home with the help of a nurse or general practitioner.
		Testosterone cypionate	50-100 mg weekly or 100-200 mg every 2 weeks	Intramuscular or subcutaneous
		Testosterone enanthate			Available in the United States as a subcutaneous autoinjector with 3 set doses.
		Testosterone esters (combination of proprionate, phenylproprionate, isocaproate, decanoate)	250 mg every 2 weeks	Intramuscular	Available in Europe.
		Testosterone gel	Varies based on formulation and strengthd	Transdermal	Caution in patients with frequent skin-to-skin contact with other people or pets.
		Testosterone patch	2-6 mg daily	Transdermal	No longer manufactured in the United States as of January 2023. Not available in all European countries.

^a While clinicians may have experience with other formulations and routes of administration (eg, pellets, buccal, compounded, oral testosterone, other antiandrogens or antiestrogens), only medications described in currently published guidelines are included (3, 7, 8). Decisions to start with or switch to specific regimens may be influenced by patient characteristics and medical history, patient or prescriber experience, cost or insurance coverage, availability of the medications, and other factors. Starting doses are often at the lower or middle range, and doses can be titrated up or down to achieve serum sex hormone levels within physiologic reference ranges that align with a patient's gender identity. Historically, recommended dosages and serum sex hormone levels have been described for binary TGD populations; more research is needed to identify goals, effectiveness, and target hormone levels for nonbinary and gender diverse groups. Considerable variability exists in prescribing practices across jurisdictions: Lower dosages may be required for safety or patient goals, while higher dosages may be necessary to achieve effectiveness if safety monitoring does not reveal any adverse effects.

^b The World Professional Association for Transgender Health and Endocrine Society include both once weekly and every 2 weeks frequencies for estradiol valerate and estradiol cypionate, while the European Society for Sexual Medicine advises 5 to 30 mg every 2 weeks for estradiol valerate only and 2 to 10 mg once weekly for estradiol cypionate only. Estradiol cypionate is available in less concentrated solution than valerate.

^c The European Society for Sexual Medicine suggests estradiol gel 1.5 mg once or twice daily.

^d Examples include 1% gel 25 to 100 mg daily, 1.6% gel 20.25 to 81 mg daily.

TGD patients with complex medical histories present to endocrinologists to initiate GAHT or for consultation after developing medical conditions that may be negatively impacted by the continuation or cessation of GAHT. Using cases from clinical practices to highlight TGD patients with complex medical histories, this Approach to the Patient manuscript summarizes how estrogen and testosterone, the mainstays of GAHT, can be managed in the setting of venous and arterial thrombosis as well as low bone density/osteoporosis.

Case 1: Venous Thromboembolism in the Setting of Feminizing GAHT

A 58-year-old transgender woman with history of hyperlipidemia and prior tobacco smoking was seen in the Endocrinology outpatient clinic after a recent pulmonary embolism. She initially received GAHT from an online source but had been under a physician's care for 4 years prior to the event. She had initiated oral estradiol but was changed to transdermal and then injectable estradiol by her physician due to her concerns about the patches being cumbersome and falling off. She had no prior personal or family history of thromboembolic events and had undergone gender-affirming orchiectomy, liposuction, and abdominoplasty without complication. At the time of her pulmonary embolism her medication regimen included estradiol valerate 2 mg weekly intramuscular injections and micronized progesterone 100 mg and atorvastatin 40 mg nightly. She presented to an emergency room with chest pain and tachycardia and was found to have small subsegmental pulmonary emboli in branches to the lingula and right lower lobe without evidence of right heart strain. She was placed on rivaroxaban and told to stop estrogen until further consultation with Endocrinology. On presentation, laboratory investigations revealed estradiol 94 pg/mL (345 pmol/L, unclear timing of the blood drawn from the last injection) and low-density lipoprotein (LDL) 79 mg/dL (2.05 mmol/L).

How Do We Approach Our Patient on Feminizing GAHT Who Had a Pulmonary Embolism?

Given concerns for adverse health outcomes and worsening gender dysphoria (if present) if feminizing GAHT is withheld, using the lowest risk formulation and route of administration based on pathophysiological and clinical studies is an option when continuing GAHT is clinically indicated. Estradiol binding to estrogen receptor-α, estrogen receptor-β, and G protein-coupled estrogen receptors in vascular endothelial and smooth muscle cells leads to vasodilatory effects, vascular tone and blood pressure control, and protection against vascular inflammation, atherogenesis, and injury (13). However, oral estrogen has been shown to affect synthesis of coagulation factors toward a prothrombotic state more than transdermal estrogen (14, 15). Specifically, oral estrogen type influences this activity; studies suggest ethinyl estradiol is of highest thrombotic risk followed by conjugated equine estrogens, 17-β estradiol, and estradiol valerate (16). Therefore, transdermal estrogen is the preferred route of therapy for people at higher risk of VTE and for older adults (eg, age >45 years per World Professional Association for Transgender Health Standards of Care 8) on feminizing GAHT (3, 7, 17, 18).

Estradiol is a well-recognized risk factor for VTE in all people, but the body of data regarding absolute risk is limited to mainly cohort studies including recent literature assessing risk among general TGD populations. The STRONG study looked at the incidence of VTE in 2842 US transgender women compared with cisgender men and women. Transgender women had significantly higher relative risk of VTE compared with cisgender men (adjusted hazard ratio [aHR] 1.9 [95% CI, 1.4-2.7]) and cisgender women (aHR 2.0 [95% CI, 1.4-2.8]), although absolute risk remained low (incidence rate, 3.1 per 1000 person-years [95% CI, 3.0-4.6]). This risk increased over time, with a 2-year risk difference of 4.1 (95% CI, 1.6-6.7) and 3.4 (95% CI, 1.1-5.6) per 1000 persons compared with cisgender men and women, respectively, while risk at 8 years was 16.7 (95% CI, 6.4-27.5) and 13.7 (95% CI, 4.1-22.7) per 1000 persons compared with cisgender men and women, respectively (9). This contrasts with earlier data and studies in cisgender women that showed most of the risk occurred earlier after initiation of estrogen, particularly in premenopausal women taking oral contraceptive pills and with thrombophilia and postmenopausal women on estrogen-based hormone replacement therapy with or without progesterone (19, 20). Transgender women also had significantly higher relative risk of ischemic stroke compared with cisgender women (aHR 1.9 [95% CI, 1.3-2.6]) but not cisgender men.

European data has also shown higher rates of VTE and ischemic stroke among transgender women compared with the general population. A review of 2517 transgender women demonstrated standardized incidence ratios for VTE (looking at observed over expected cases) of 5.52 compared with cisgender women and 4.55 compared with cisgender men (10). The standardized incidence ratios decreased slightly to 3.92 and 3.39, respectively, in a subgroup analysis excluding transgender women who initiated GAHT before 2001 (after which ethinyl estradiol was no longer used). The standardized incidence ratios for ischemic stroke in this cohort were 2.42 compared with cisgender women and 1.80 compared with cisgender men. However, subgroup analyses by age or smoking status have not yet been performed in transgender women.

It is important to note that the large retrospective studies on VTE and ischemic stroke could not account for route of administration, dose, or type of estrogen prescribed. In addition, these studies did not account for preexisting morbidities that increase VTE and ischemic stroke risk or their risk factors. Despite the increases in relative risk, the absolute risk of VTE among transgender women is low, with studies estimating 2.3 per 1000 patient years (21). This is higher than the general population (estimated 1.0-1.8 per 1000 patient years), premenopausal cisgender women on oral contraceptive pills (0.8-1.0 per 1000 patient years), and postmenopausal cisgender women aged 40-50 years (0.5-1.2 per 1000 patient years) but lower than the VTE risk during the postpartum period in cisgender women (4.2 per 1000 patient years) and postmenopausal cisgender women (inclusive of women with and without hormone replacement therapy) aged 60-70 years (3.0-7.0 per 1000 patient years) (22-26). A recent systematic review and meta-regression analysis of 18 studies (n = 11 542 study participants) calculated a pooled VTE prevalence in assigned male at birth individuals on feminizing GAHT of 2% (27). Pooled prevalence for age <37.5 years and mean duration of GAHT <53 months was 0%, though the study did not take route of administration into account. However, the individual risk for GAHT-related VTE of the patient in Case 1 is estimated to be higher than that of the general population of transgender women, due to her history of hyperlipidemia, tobacco smoking, and prior VTE while taking GAHT.

The patient in Case 1 was also on micronized progesterone at the time of her VTE and wondered what role, if any, this may have played in the event. Studies of cisgender women have suggested that micronized progesterone may have fewer metabolic adverse effects than medroxyprogesterone (3, 28). Given the absence of evidence supporting a benefit of progestins in feminizing GAHT, guidelines do not currently recommend their routine use in GAHT regimens (3, 7, 8, 29). A recent systematic review examined the use of progesterone as part of GAHT (30). Ten studies met criteria for inclusion. One of the observational studies including 2509 transgender women reported a higher incidence of progesterone prescription in the 19 subjects who developed VTE (26% with a progesterone prescription vs 14% in those without) (31). This association persisted (odds ratio [OR] 2.95 [95% CI 1.02-8.57], P = .04) even after adjustment for race, insurance, HIV status, age, and body mass index (BMI), but the authors note the limitations of a small number of events and wide confidence intervals.

Hematology colleagues with expertise in transgender health should be consulted in cases of VTE or ischemic stroke as they can help assess TGD patients for underlying thrombophilia and provide patients with the risk and benefits and cost of various anticoagulation preparations, if indicated. Studies suggest that estrogen use in people with a Factor V Leiden mutation can have increased risk of VTE by 15 to 35 times (14). However, preemptive screening for thrombophilia is not cost-effective and should not be done unless there is a clinical concern based on personal or family history reported by the patient (16). In one study, partial thrombophilia screening uncovered low incidence among 162 transgender women (8.0%) and 89 transgender men (5.6%), with no VTE cases during a mean of 49.6 (SD 33.7) months of GAHT follow-up, including among 18 transgender women with activated protein C resistance treated with transdermal estrogen (of those, 1 person had history of VTE and was also taking long-term anticoagulation) (32).

A consultation with Hematology can be considered prior to GAHT initiation or continuation in patients with strong risk factors or a family history of thromboembolic events to determine if testing and/or anticoagulation should be utilized (33). In patients with prothrombotic mutations, nonprovoked hypercoagulable states, and/or prior VTE while on estrogen (including using transdermal estrogen), providers should discuss the risks and benefits of switching to or resuming transdermal estrogen after acute VTE management and consider long-term anticoagulation or antiplatelet therapy, if indicated (34). It has been proposed that any type of estrogen can be used in patients with thrombosis history if they continue taking anticoagulation, given the data showing anticoagulation prevents new thrombosis (ie, no increased risk of VTE) (32, 35, 36). Hematology colleagues can also advise patients and clinicians regarding preparations for surgery and how to adjust medications during the peri- and postoperative period.

How Do We Advise the Patient in Case 1?

The risks of VTE and ischemic stroke increase in the presence of other cardiovascular risk factors including increasing age, tobacco use, obesity, hypertension, and adverse lipid profiles (16). This patient in Case 1 did not have hypertension (although she used spironolactone as part of GAHT), had a BMI of 26 kg/m², was already on statin therapy, and had a prior but not current history of tobacco use. It is important to work with patients to optimize modifiable factors that contribute to increased risk of thromboembolic (and cardiometabolic) disease. This patient strongly desired to continue GAHT and had already failed transdermal estradiol due to poor absorption and skin irritation. After discussion about potential supraphysiologic levels of estradiol and associated risks, she was continued on low-dose weekly injectable estrogen along with anticoagulation. The thought in the approach to this patient was to avoid the hepatic first pass metabolism with oral estradiol and injections were her desired route. Her follow-up estradiol levels have all been <100 pg/mL (<367.10 pmol/L), and she has been satisfied with the physical and mental health benefits of continuing estrogen. While guidelines recommend achieving serum estradiol levels in the 100 to 200 pg/mL range with feminizing GAHT, they do not provide “goal” estradiol levels in high-risk patients (3, 7, 8). An approach to managing high-risk patients is to avoid supraphysiologic serum estradiol levels, which is consistent with peer-reviewed guidelines.

Case 2: Ischemic Stroke in the Setting of Masculinizing GAHT

A 63-year-old transgender man with a history of hypertension, hyperlipidemia, prediabetes, overweight BMI, and depression was referred to Endocrinology outpatient clinic by his primary care physician for management of GAHT following an ischemic stroke the year prior. The stroke was associated with hemiplegia and aphasia, consistent with involvement of the left middle cerebral artery. At the time of the stroke, he had been on injectable testosterone for 10 years and was injecting testosterone cypionate 100 mg intramuscularly weekly. A few months before the stroke, the total testosterone level was 928 ng/dL (32.20 nmol/L, unclear timing of the venipuncture since the last injection) and his hematocrit was 47%. The patient underwent bilateral mastectomy 10 years ago but had had no other surgeries. He had a 10-year cigarette smoking history but quit after the stroke. The stroke treatment team discontinued the testosterone injections and suggested outpatient Endocrinology consultation.

How Do We Approach Our Patient Who Had an Ischemic Stroke on Masculinizing GAHT?

Concerns have been raised that testosterone therapy in cisgender men with hypogonadism may be associated with increased risk of cardiovascular events although recent data have been more reassuring, including a randomized placebo-controlled trial in cisgender men with preexisting or a high risk of cardiovascular disease (37-39). Smaller studies have not consistently revealed higher risk of cardiovascular disease among the general population of transgender men taking testosterone compared with transgender women and cisgender men and women (9, 10, 40-42). Additionally, testosterone replacement for transgender men, particularly when physiologic testosterone levels are achieved, has not been associated with an increased risk of thrombosis (16). Neither intramuscular nor transdermal testosterone affect coagulation factors toward a prothrombotic state among transgender men (15).

At a physiological level, testosterone binding to androgen receptors in vascular endothelial and smooth muscle cells leads to both vasodilatory and vasoconstrictive effects, cardioprotection following ischemia via upregulation of cardiac α1-andrenoceptor but may also contribute to vascular inflammation and oxidative stress (13). These paradoxical actions do not seem to favor harm in the TGD population. In the STRONG cohort referenced above, 2118 US transgender men were compared with cisgender men and women and were not found to have significantly higher risk of VTE or ischemic stroke compared to cisgender populations (9). Similarly, European data have not revealed a higher risk of VTE or ischemic stroke compared to cisgender populations (10).

While testosterone has not been associated with procoagulant changes among transgender men (15), hematocrit will increase in any person on testosterone due to the effect on erythropoiesis, regardless of route of administration, although the clinical significance remains to be determined (14, 43-45). Data from a large Dutch cohort of transgender men (n = 1073) taking testosterone for 20 years revealed that the prevalence for hematocrit >50%, > 52%, and >54% was 11%, 3.7%, and 0.5%, respectively; the largest increase in hematocrit from baseline occurred during the first year on testosterone and the probability of developing erythrocytosis persisted beyond 10 years (46). An Australian cohort of 180 transgender men taking testosterone a median of 37.7 months (and 27% were smokers) had polycythemia prevalence of about 25% with intramuscular testosterone enanthate and 15% with intramuscular testosterone undecanoate, while no one on transdermal testosterone had polycythemia (47).

How Do We Advise the Patient in Case 2?

After being off testosterone for a year after his stroke, the patient in Case 2 developed symptoms of fatigue, generalized pain, and decreased facial hair. He worked with Neurology on behavioral modification and medication compliance for secondary stroke prevention (ie, antihypertensives, statin, antiplatelet treatment) and desired to reinitiate testosterone therapy. The approach in this case involved discussing the available routes of testosterone administration and his preferred route, balancing the impact of continuing to withhold testosterone therapy on his mental health with the physical benefits and possible adverse events related to taking exogenous testosterone. Although transdermal testosterone, via patch or gel, can increase hematocrit as mentioned above, the clinical judgment was made to change him to testosterone gel, although we lack studies evaluating the benefits of switching a high-risk patient from intramuscular injection to transdermal testosterone (14).

Case 3: Osteoporosis in the Setting of Feminizing GAHT

A 42-year-old transgender woman self-referred to Endocrinology outpatient clinic for joint pains and hot flashes. She started feminizing GAHT as a young adult but used this intermittently both prior to and after her orchiectomy and vaginoplasty 5 years earlier. When last seen 4 years ago, she was taking a self-chosen combination of transdermal and injectable estrogens, and her estradiol level was measured at 480 pg/mL (1762.24 pmol/L). She reported choosing to stop her GAHT several months later, both because she had met her gender affirmation goals and due to concerns about weight gain and emotional lability. At her return visit, she reported ongoing hot flashes and painful and stiff joints at her fingers, knees, and ankles. She reported infrequent cigarette use. Her only other medication was a salmeterol/fluticasone inhaler. Dual-energy x-ray absorptiometry (DXA) was ordered due to her longstanding nonadherence with sex hormone use after gonadectomy. It revealed bone densitometry at the lumbar spine of 0.725 g/cm² (Z-score −3.2, −2.6, T-score −3.3, −2.9), femoral neck 0.653 g/cm² (Z-score −1.5, −1.4, T-score −2.0, −1.8) and total hip 0.968 g/cm² (Z-score −0.2, + 0.4, T-score −0.4, + 0.2) according to cisgender male and female reference values, respectively. She had no prior DXA, did not report any previous fractures, had a normal calcium level (9.36 mg/dL, 2.34 mmol/L), and her 25-hydroxy vitamin D level was 22.1 ng/mL (55.16 nmol/L).

What Is the Role of Sex Steroids in BMD Acquisition and Maintenance?

Before puberty, linear and radial bone growth are mainly dependent on growth hormone and IGF-1 (48). Skeletal differences between males and females arise during puberty when production of sex steroids begin to diverge, and males will typically achieve a higher peak bone mass compared to females (see Fig. 1) (50). Cisgender male puberty is characterized by more periosteal cortical expansion compared to endosteal cortical resorption, thus resulting in wider bones and subsequently higher bone mass. Cisgender female puberty is characterized by less periosteal expansion compared to males, but more endosteal growth, leading to narrowing of the medullary cavity, although cortical density is comparable in females and males (51). The differential roles of estrogens and androgens on the bone in male puberty have not fully been elucidated. Testosterone may act directly on the bone via androgen receptor activation, as well as indirectly by increasing muscle mass and thereby increased mechanical loading (52, 53). Testosterone aromatization to estradiol plays a key role in attaining peak bone mass during puberty. Estrogen inhibits sclerostin, decreases osteoblast and osteocyte apoptosis, suppresses RANK-L, and induces apoptosis in osteoclasts, as well as promoting bone formation by inhibiting osteoblast and osteocyte apoptosis (48).

Figure 1.

Bone changes. Bone densitometry and geometry changes in transgender women (above) and transgender men (below) before and 10 years after GAHT as compared with postpubertal cisgender men and cisgender women, respectively. Transgender women, before initiation of GAHT, have lower areal BMD (aBMD) (as indicated by a lighter shade of gray), smaller periosteal circumference, and smaller cortical bone area and thickness as compared with cisgender men. An increase in aBMD is observed over the first 2 years (not shown) but, by 10 years, a decline in aBMD toward baseline may occur. Transgender men, before starting GAHT, have similar aBMD and bone geometry as compared with cisgender women (as indicated by the same shade of gray). Up to 10 years after GAHT initiation, aBMD continues to increase in transgender men who initiated GAHT ≥ 40 years old; aBMD remains unchanged in transgender men who initiated GAHT at age < 40 years (not shown separately). There have been studies reporting larger cortical size in transgender men on testosterone although data are limited. For transgender men and women, there remains a lack of qCT data at 10 years; the horizontal lines and or dots represent speculation. Fracture Risk. Based on Wiepjes et al (49), transgender women aged ≥ 50 years with median duration of GAHT 19 (IQR, 11-29) years had higher fracture risk compared to age-matched cisgender men (as indicated by the larger fracture in the bone in the former); however, they had similar fracture risk compared with cisgender women. Transgender women aged < 50 years with median duration of GAHT 8 (IQR, 3-16) years tended to have a higher fracture risk compared to age-matched cisgender women but similar fracture risk compared to cisgender men. Transgender men with mean age 40 (SD 14) years and median duration of GAHT 9 (IQR, 2-22) years did not have an increased fracture risk compared to age-matched cisgender men or women (as indicated by the smaller fracture in the bone in the former; fracture risk in transgender men was not stratified by age group). Abbreviations: GAHT, gender-affirming hormone therapy; IQR, interquartile range.

Both scientific studies and examples in nature have shown that estradiol is the key to bone health maintenance. When cisgender females go through menopause, estrogen deficiency leads to increased osteoclastic bone resorption and decreased bone mass. Reducing estrogen levels by adding aromatase inhibitors to the treatment of postmenopausal women in oncological settings results in significant bone loss, demonstrating the impact of locally derived estradiol on the maintenance of skeletal health. The importance of estrogen for bone health in cisgender males can be deduced from studies using aromatase inhibitors or androgen deprivation therapy. Finkelstein et al described a decrease in BMD in males who were first administered a gonadotropin-releasing hormone (GnRH) agonist and subsequently administered testosterone plus an aromatase inhibitor (54). However, no significant change in BMD was observed in people receiving a GnRH agonist and testosterone, without an aromatase inhibitor. Estrogen deficiency in early life (eg, aromatase deficiency, estrogen alpha receptor mutation) inhibits the growth spurt and subsequent fusion (55, 56) regardless of sex assigned at birth. Additionally, estrogen treatment in men with aromatase deficiency leads to increased longitudinal bone growth and cross-sectional area and cortical thickness (57), whereas these effects are not seen in the setting of estrogen receptor-α mutations.

How Does Feminizing GAHT Affect Bone Health in TGD Patients and How Do We Approach Our Patient on Feminizing GAHT With Low BMD?

Many studies have shown bone density to be low in transgender women even prior to the initiation of GAHT: 14% to 18% with T-scores ≤ −2.5 and up to 22% with Z-scores < −2.0 (58) (see Fig. 1). This has been attributed to vitamin D deficiency and decreased physical activity with lower muscle mass and therefore less mechanical loading on the bone (59). While on feminizing GAHT, studies have shown that transgender women with the lowest quartile estradiol level (eg, 115 pmol/L, 31.3 pg/mL) had the highest bone turnover markers compared to the higher quartile estradiol levels, and lower estradiol levels (ie, < 182 pmol/L, 49.6 pg/mL) have been associated with significant decreases in BMD compared to higher estradiol levels (12, 60).

However, exogenous estrogen is beneficial for bone health, even in the face of suppressed testosterone levels. A 2019 meta-analysis of longitudinal studies confirmed a small increase in spine areal BMD during the first year of feminizing GAHT but suggested unchanged femoral neck and total hip areal BMD (61). However, studies still report a prevalence of up to 40% of low bone mass in transgender women after long-term GAHT (58).

In the largest retrospective cohort to date, the 3-year fracture risk has been described as higher in transgender women over age 50 years (4.4%) using long-term GAHT, compared with age-matched cisgender men (2.4%), but similar to age-matched cisgender women (4.2%), even if decreasing or discontinuation of estradiol supplementation at that age was not part of the treatment protocol (see Fig. 1). Relatively more hip, spine, forearm, and humerus fractures were seen in transgender women compared with cisgender men (49). In transgender women under the age of 50 years, fracture risk was comparable to cisgender men, but higher than in cisgender women.

Cross-sectional bone microarchitecture data measured by high-resolution peripheral quantitative computed tomography (HRpQCT) revealed that transgender women on established GAHT had compromised bone microarchitecture relative to cisgender men with lower total volumetric BMD (vBMD), higher cortical porosity, and lower trabecular density (62). Contrary to the data from areal bone densitometry, fracture data and HRpQCT data suggest that bone may potentially be compromised in transgender women on GAHT, but it is difficult to distinguish whether this could be attributed to GAHT or was already present before the initiation of GAHT.

In Case 3, screening for osteoporosis seemed appropriate to do since the patient had a prolonged period of hypogonadism and prior tobacco use (63). Baseline data before GAHT were not available. There are no transgender-specific guidelines for osteoporosis treatment. The Endocrine Society and International Society for Clinical Densitometry (ISCD) do not advise routine assessment of BMD in transgender people; performing DXA is only advised in the event of risk factors for osteoporosis (11, 58). Suboptimal compliance with hormone therapy, the use of aromatase inhibitors, or stopping estrogen treatment after gonadectomy are indications for osteoporosis screening, regardless of age (11). As the ISCD recommends comparing bone densitometry to Z-scores for the affirmed gender, one may choose to follow treatment guidelines for the affirmed gender or choose to follow the most cautious guidelines.

General bone health measures such as adequate calcium and vitamin D intake, a healthy lifestyle with adequate physical activity, and optimizing treatment compliance and ensuring adequate exposure to sex steroids is recommended (11). If a person does not want to be on GAHT or if the contraindications present too much risk, pharmacological treatment for osteoporosis could be discussed in addition to the general bone health measures mentioned above. For an overview, we refer to the Endocrine Society Clinical Practice Guidelines for osteoporosis (64, 65).

How Do We Advise Our Patient With Low BMD Who Has Stopped GAHT?

After acknowledging the patient's gender affirmation goals, addressing concerns for weight gain and emotional lability, and discussing ongoing hot flashes and risk to her bones while being off estrogen after orchiectomy, we advised the patient to reinitiate oral estradiol 2 mg daily (transdermal estradiol was not well tolerated) with an increase up to 4 mg daily if tolerated well without adverse effects, along with adding calcium and vitamin D (1000 mg/800 IU) supplement to the treatment. In addition, the importance of general bone health measures (eg, exercise, tobacco cessation, dietary calcium) was explained. Four months later, patient adherence was good and she reported feeling well, stable emotionally, with quick disappearance of the joint pains. Based on her age and lack of fracture history, additional treatment was not considered necessary. The plan was to repeat the DXA in 2 years.

Case 4: Bone Density in the Setting of Masculinizing GAHT

A 46-year-old transgender man with a history of osteoporosis, chronic kidney disease due to overuse of nonsteroidal anti-inflammatory agents and dehydration, hypercalcemia in the setting of acute kidney insufficiency and elevated 25-hydroxy vitamin D levels, anorexia nervosa, depression, substance use and suicide attempt, was seen in the Endocrinology outpatient clinic. He started GAHT at age 35 years and underwent gonadectomy and mastectomy. He was taking testosterone undecanoate 1000 mg intramuscularly once every 11 weeks, omeprazole 40 mg twice daily, lormetazepam 2 mg once daily, and paracetamol 1 g as needed for headaches. Psychological distress often resulted in worsening gastrointestinal reflux and increased antacid use. He decreased his alcohol consumption after years of heavy drinking. Laboratory investigation showed a serum creatinine level of 4.3 mg/dL (380.12 umol/L) and estimated glomerular filtration rate < 15 mL/min/1.73 m², both worse compared to the previous month. His testosterone level was 1000 ng/dL (34.70 nmol/L), measured soon after injection, with hematocrit of 40.2%. Given his kidney disease and anorexia nervosa, a baseline DXA was obtained and revealed lowest Z-scores of −2.2 and −2.0 and T-scores of −3.7 and −3.9 at the total hip (compared to cisgender male and female reference values, respectively). He was concerned whether his GAHT would have a negative impact on his bone health. As shown in Fig. 2, BMD increased over the first 5 years after the initiation of GAHT (at age 35 years old), as the patient gradually gained weight, possibly related to improved dietary habits and his starting of an exercise regimen. However, BMD decreased slightly over the last few years (between 41 to 46 years old), which may be explained by tertiary hyperparathyroidism due to chronic renal failure and ongoing alcohol consumption. He was advised to stop antacids and continue to decrease the amount of alcohol consumed.

Figure 2.

Case 4 patient's bone densitometry as measured by hologic discovery a, over time. Bone densitometry was performed before the initiation of GAHT (2011, 34 years old), after 1 year of GAHT (2012, 35 years old), 5 years of GAHT (2016, 40 years old), 7 years of GAHT (2018, 42 years old), and 10 years of GAHT (2022, age 45). Abbreviations: AP, anterio-posterior; BMD, bone mineral density (g/cm²); L1-L4, lumbar vertebrae 1 to 4.

How Does Masculinizing GAHT Affect Bone Health in TGD Patients?

Most studies on BMD and fracture risk confirm that GAHT does not impact bone health in transgender men (except for 2 studies reporting small increases in spine, hip, or femoral neck BMD over the first 2 years of GAHT and one study reporting a decrease in spine BMD over 4.5 years of GAHT) (see Fig. 1) (58). Case 4 showed an increase in BMD after starting GAHT. Some of this may be attributed to an increase in lean mass the patient experienced after initiating testosterone. A meta-analysis including 10 studies revealed an increase in body weight (+1.7 kg [95% CI 0.7; 2.7]) and lean body mass (+3.9 kg, [95% CI 3.2; 4.5]) and a decrease of body fat (−2.6 kg [95% CI −3.9; −1.4]) (66) Additionally, an earlier meta-analysis of 44 studies between 1989 and 2013 revealed that lean mass has a greater effect on bone density compared with fat mass in presumed cisgender men and women (67). Previous studies in TGD adults examining BMD and grip strength and/or lean body mass revealed no correlation between them (68, 69). These results may be due to a relatively short duration of testosterone therapy, as it may require longer-term testosterone therapy to achieve the same levels of muscle mass/strength and bone mass as cisgender men.

Research on bone microarchitecture in transgender people is limited by the short follow-up duration and smaller sample size, which may lead to underpowered results. In older transgender men (≥ 40 years old), GAHT may improve subperiosteal width, endocortical diameter, and trabecular bone score, compared to not taking GAHT (70). In younger transgender men (< 40 years old), trabecular bone scores tend to be lower in those using GAHT compared with those not using GAHT (70). In contrast, a prospective cohort study of transgender men of all ages showed a small increase in trabecular vBMD (measured by quantitative computed tomography, qCT) at the distal radius and another cross-sectional study revealed higher vBMD and trabecular thickness at the distal tibia in transgender men compared with cisgender women (53). No significant difference was observed regarding porosity of the cortex, trabecular bone volume/tissue volume, trabecular number, and trabecular separation (62). In transgender men, fracture risk has been shown to be comparable to cisgender women but lower than in cisgender men (49). Meriggiola et al reported no changes in aBMD in transgender men during the first year of GAHT, but a decrease in those who were treated with testosterone in combination with an aromatase inhibitor (71). Overall, these data suggest no increased risk for abnormal bone health in transgender men after starting GAHT, and the estradiol level present from aromatization may be adequate for skeletal protection.

How Do We Advise the Patient With Multiple Medical Comorbidities and Baseline Low BMD on Masculinizing GAHT?

With the available data on bone health noted above, there is no contraindication to starting GAHT in people with low T-scores or Z-scores. If osteopenia/osteoporosis are diagnosed, a healthy lifestyle with weight-bearing physical activity, adequate calcium and vitamin D intake is advised with pharmacologic therapy, if deemed appropriate after assessing fracture risk. The importance of compliance with GAHT for bone health should be emphasized, particularly after gonadectomy (11).

The patient in Case 4 had a marked improvement in BMD after starting GAHT. Qualitative research in transmasculine people have reported high attention to nutrition and exercise, which may be used as tools to achieve a desired body size and shape (72). However, co-occurring psychological difficulties in Case 4 led to medication and substance abuse, with subsequent kidney dysfunction and loss of BMD. The impact of his psychological difficulties was discussed with him, and more frequent follow-up with a mental health professional was offered. The case was regularly discussed during multidisciplinary team meetings, recommendations from mental health colleagues were shared with the patient, and improvements were seen in his overall and mental health. This highlights the importance of regular follow-up visits consisting of mental health questions, routine health questions (eg, tobacco and substance use, symptoms of depression, risk behavior), physical examination (including weight and blood pressure), laboratory monitoring, and multidisciplinary discussions. The Endocrine Society Clinical Practice Guidelines suggest follow-up visits every 3 months during the first year of GAHT and then once or twice yearly (7). As mentioned above, screening for bone loss is advised in people stopping testosterone treatment after gonadectomy, people who are not compliant with GAHT or people presenting with other risks for bone loss.

Conclusions

The clinical management considerations presented in the cases above demonstrate the need for a case-by-case assessment of prescribing GAHT in the setting of thromboembolic and cardiovascular risk and low bone density/osteoporosis. The informed shared decision-making process between clinician and patient promotes the balancing of GAHT benefits on mental and physical health with potential medical risks. In most situations, patients will generally choose GAHT initiation or continuation, and it is the role of the clinician to approach these patients in partnership to mitigate risks as best as possible using available guidelines and supporting literature.

Abbreviations

aHR

adjusted hazard ratio

BMD

bone mineral density

BMI

body mass index

DXA

dual-energy x-ray absorptiometry

GAHT

gender-affirming hormone therapy

HRpQCT

high-resolution peripheral quantitative computed tomography

ISCD

International Society for Clinical Densitometry

TGD

transgender and gender diverse

vBMD

volumetric bone mineral density

VTE

venous thromboembolism

Funding

This work was supported by the National Institutes of Health and University of Colorado Building Interdisciplinary Research Careers in Women's Health program [NIH K12 HD057022] (S.J.I.) and by Fonds voor Innovatie en Klinisch Onderzoek (FIKO) Ghent University Hospital (G.T.).

Disclosures

The authors have no conflicts of interest to disclose.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.