Relationship Between Serum Estradiol Concentrations and Clinical Outcomes in Transgender Individuals Undergoing Feminizing Hormone Therapy: A Narrative Review

Abstract

Transgender, including gender diverse and nonbinary, individuals are treated with estradiol with or without antiandrogen to align their physical appearance with their gender identity, improve mental health and quality of life. Consensus guidelines give target ranges for serum estradiol concentration based on premenopausal female reference ranges. However, limited studies have evaluated the relationship between serum estradiol concentrations and clinical outcomes in transgender individuals undergoing feminizing hormone therapy. The available evidence has not found that higher serum estradiol concentrations, together with suppressed testosterone, enhance breast development, or produce more feminine changes to body composition. However, ensuring testosterone suppression appears to be an important factor to maximize these physical changes. Higher serum estradiol concentrations have been associated with higher areal bone mineral density. Although the resultant long-term clinical implications are yet to be determined, this could be a consideration for individuals with low bone mass. The precise serum estradiol concentration that results in adequate feminization without increasing the risk of complications (thromboembolic disease, cholelithiasis) remains unknown. Further prospective trials are required.

Introduction

Transgender, including gender diverse and nonbinary, individuals who seek feminization are often treated with estradiol with or without antiandrogen to increase serum estradiol concentration and decrease serum testosterone concentration into the reference ranges for premenopausal women. Physical changes, including softening of skin, a decrease in facial and body hair, breast development, and feminine changes to body composition, develop within months although maximal effects may take 2–3 years.¹

Consensus guidelines give recommendations for serum estradiol concentration targets to permit titration of estradiol therapy.^1–3 The 2017 Endocrine Society Clinical Practice Guidelines recommend maintenance of serum estradiol concentrations between, and not exceed, 100–200 pg/mL (367–734 pmol/L) and testosterone concentrations <50 ng/dL (1.7 nmol/L).¹ These values are derived from sex steroid concentrations in premenopausal women and represent a surrogate target to enable suppression of testosterone while minimizing supraphysiological estradiol concentrations. However, no study has been performed to establish optimal estradiol concentrations to promote feminization in transgender individuals undergoing feminizing hormone therapy. When considering optimal serum estradiol concentrations, adequate feminization must be weighed against potential risks, given that escalating oral estradiol dose has been associated with an increased thromboembolic risk in the menopausal hormone therapy literature.⁴

Given the uncertainty regarding estradiol concentration targets in transgender individuals undergoing feminizing hormone therapy, we undertook a narrative review to establish the relationship between serum estradiol concentrations and clinical outcomes. Herein, we present a summary of the literature regarding the relationship between serum estradiol concentrations and breast development, body composition, and bone health in transgender individuals. We also consider risks associated with escalating estradiol dose and serum estradiol concentration. MEDLINE, EMBASE, and PsycInfo were searched using MeSH terms and text words for transgender, estradiol, breast, body composition, and bone. No randomized controlled trials were found, so data have been obtained from prospective cohort, retrospective cohort, and cross-sectional studies. We also searched the references listed in relevant publications.

Breast development

Breast development is a key clinical outcome for many transgender individuals. However, breast development is often modest,^5–8 and many individuals seek surgical breast augmentation.^6,9 Two studies have reported the influence of serum estradiol concentration on breast development in transgender individuals.

de Blok et al. performed a 12-month prospective study as part of the European Network for the Investigation of Gender Incongruence (ENIGI) to evaluate breast development in 229 transgender individuals newly commencing gender-affirming hormone therapy (GAHT).⁵ Breast development, as measured by changes in breast-chest circumference and cup size, was modest and predominantly occurred within the first 6 months of GAHT. Mean breast-chest difference increased to 7.9±3.1 cm after 12 months, with nearly half of participants having less than AAA cup size.

Serum estradiol concentration, taken as an average of two values performed at months 3 and 12, did not predict breast development after 12 months of GAHT (Table 1). Only the highest quartile (mean serum estradiol concentration 452 pmol/L in Amsterdam and 567 pmol/L in Ghent) reached serum estradiol concentrations recommended in Endocrine Society guidelines. Testosterone was suppressed in 92% of individuals, so further analyses evaluating the influence of serum testosterone concentration could not be performed.

Table 1.

Influence of Serum Estradiol Concentration on Clinical Outcomes in Transgender Individuals Undergoing Feminizing Hormone Therapy

Reference	Country	Study type	Duration	Number of individuals	Results
Breast development
de Blok et al.5	Netherlands, Belgium, and Italy (ENIGI)	Prospective, cohort	12 months	229	Serum estradiol concentration did not predict breast development (first quartile, 3.6 cm [95% CI: 2.7–4.5], second quartile, 3.2 cm [95% CI: 2.3–4.2], third quartile, 4.4 cm [95% CI: 3.5–5.3], and fourth quartile, 3.6 cm [95% CI: 2.7–4.5])
Meyer et al.10	Germany	Retrospective, cohort	Up to 3–4 years	155	Estradiol concentration had no significant influence on breast development (Spearman correlation [ρ]=−0.117, p=0.316)
Body composition
Klaver et al.14	Netherlands and Belgium (ENIGI)	Prospective, cohort	12 months	179	No association between mean serum estradiol concentration and body fat or lean body mass (p=not reported)
Klaver et al.26	Netherlands	Retrospective, cohort	Mean 7.5 years	71	Serum estradiol concentration increased from 25 to 121 pmol/L after commencement of estradiol. Percentage of lean body mass increased 3%,1,5 percentage of total fat mass decreased −3% (−5,−1), and WHR decreased −0.02 (−0.04, 0.01) after commencement of estradiol
Bone health
Lapauw et al.36	Belgium	Cross-sectional	N/A	23	Serum sex steroid concentrations not associated with pQCT parameters (p>0.18)
Van Caenegem et al.20	Belgium (ENIGI)	Prospective, cohort	12 months	49	Serum estradiol concentration not associated with changes in BMD or pQCT parameters (data not shown)
Wiepjes et al.30	Netherlands and Belgium (ENIGI)	Prospective, cohort	12 months	231	Amsterdam: Estradiol concentration correlated with LS (per 100 pmol/L: +0.95%, 95% CI: 0.34 to 1.56, p=0.003), TH (per 100 pmol/L: +0.48%, 95% CI: 0.04 to 0.93, p=0.034), and FN (per 100 pmol/L: +0.83%, 95% CI: 0.31 to 1.36, p=0.002) BMD change
					Ghent: Estradiol concentrations correlated with LS (per 100 pmol/L: +0.87%, 95% CI: 0.27 to 1.47, p=0.005), but not with TH (per 100 pmol/L: +0.40%, 95% CI: −0.12 to 0.92, p.0=126) or FN (per 100 pmol/L: +0.09%, 95% CI: −0.67 to 0.85, p=0.814) BMD change
Wiepjes et al.34	Netherlands (ACOG)	Retrospective, cohort	Up to 10 years	711	Higher estradiol tertiles were associated with higher LS BMD than lower estradiol tertiles:
					Second tertile vs. first tertile +0.033 (0.006 – 0.059)
					Third tertile vs. first tertile +0.076 (0.050 – 0.103)
					Third tertile vs. second tertile +0.044 (0.018 – 0.070)
Vlot et al.33	Netherlands and Belgium (ENIGI)	Prospective, cohort	12 months	121	Sclerostin decreased in all but the lowest estradiol quartile (p≤0.05 vs. other quartiles)
					No significant differences were seen with CTx, P1NP or ALP between estradiol quartiles
Wiepjes et al.35	Netherlands (ACOG)	Retrospective, cohort	Median 8 years (aged <50), 19 years (aged ≥50)	2023 (1089 aged <50, 934 aged ≥50)	On univariable analyses, no association was found between estradiol concentration and fracture risk (per 10 pmol/L: OR=0.99, 95% CI: 0.97–1.02)

ACOG, Amsterdam Cohort of Gender Dysphoria; ALP, alkaline phosphatase; BMD, bone mineral density; CI, confidence interval; CTx, C-terminal telopeptide of type 1 collagen; ENIGI, European Network for the Investigation of Gender Incongruence; FN, femoral neck; LS, lumbar spine; OR, odds ratio; P1NP, procollagen type 1 N propeptide; pQCT, peripheral quantitative computed tomography; TH, total hip; WHR, waist-hip ratio.

Breast development was also analyzed by Meyer et al., who reported a retrospective cohort study involving 155 transgender individuals.¹⁰ Similar to the previous study, the authors also reported breast development by breast-chest difference but also by Tanner stage. Median breast-chest difference measured 8.5±5 cm and Tanner stage 4±1 at 3–4 years with estradiol concentration 342±382 pmol/L. Serum estradiol concentration had no significant influence on breast development (Spearman correlation [ρ]=−0.117, p=0.316) (Table 1). Owing to its retrospective nature, there was greater variability in the hormone regimen used, and 17% of individuals were not treated with antiandrogen therapy, which permitted further analysis of the influence of serum testosterone concentration. Both serum testosterone concentration and free androgen index (FAI) were significantly negatively correlated with the level of Tanner stage (testosterone: ρ=−0.398, p<0.001; FAI: ρ=0.346, p=0.004).

Older studies have also reported the influence of estradiol dose on breast development. One cross-sectional analysis of 38 transgender individuals reported higher breast hemicircumference with higher estradiol doses,¹¹ whereas a subsequent longitudinal study enrolling 60 individuals did not.¹² It should be noted that these studies utilized ethinyl estradiol and conjugated estrogens, which are no longer recommended as part of feminizing hormone regimens given inability to measure serum estradiol concentrations and increased risk of thromboembolic disease.^1,13

In summary, serum estradiol concentration has not been found to be associated with breast-chest difference or Tanner stage in transgender individuals. However, ensuring suppression of testosterone might be an important factor to promote maximal breast development. Limited data are available examining bra cup size or development of the nipple-areolar complex and this requires further study.

Body composition

Several prospective studies have examined changes in body composition after initiation of GAHT in transgender individuals.^14–26 Typical changes include a reduction in lean mass and increase in fat mass, particularly gynoid fat.^27,28 However, the current literature is limited by a lack of controlled studies and only one study has examined the relationship between changes in body composition and serum estradiol concentration.

Klaver et al. reported changes in body composition using whole-body dual-energy x-ray absorptiometry (DXA) after commencement of GAHT in a 12-month prospective observational study that enrolled 179 transgender adults.¹⁴ Resultant changes in body composition included increased body fat, more predominant at the gynoid region, and a decrease in waist-to-hip ratio. Mean serum estradiol concentration measured at months 3 and 12 was not found to be associated with changes in body fat or lean body mass.¹⁴

The use of gonadotropin-releasing hormone agonists (GnRHa) with subsequent commencement of estradiol in transgender adolescents allows further evaluation of changes directly attributable to estradiol. Klaver et al. reported changes in body composition in a retrospective study involving 71 transgender adolescents.²⁶ GnRHa were used to achieve puberty suppression, with GAHT commenced from 16 years of age. At 22 years of age, body composition parameters, including waist-hip ratio, total body fat, and lean body mass, were more closely aligned with that of cisgender women.

Larger changes in total body fat (+6% [4, 7] vs. +3% [1, 5]) and lean body mass (−6% [−7, −4] vs. −3% [−5, −1]) were seen during GnRHa monotherapy than after commencement of estradiol. This could suggest that suppression of testosterone has a greater impact on body composition than estradiol. However, it remains unclear if these changes are directly attributable to testosterone suppression or the sex steroid-deficient state achieved during GnRHa therapy.

In summary, serum estradiol concentration has not been found to be associated with changes in body composition in transgender adults. Although changes in body composition are seen after sequential commencement of estradiol in transgender adolescents, these changes were smaller than those seen after the initial commencement of GnRHa monotherapy.

Bone health

Despite reports of low bone mass in transgender individuals before initiation of GAHT,^20,29 prospective studies^20,22,30,31 and a meta-analysis³² have demonstrated improvements in areal bone mineral density (BMD) after commencement of GAHT. Several studies have examined the association between serum estradiol concentration and bone outcomes in transgender individuals.

Two prospective studies in the ENIGI cohort have evaluated changes in areal BMD and bone turnover markers in transgender individuals for the first 12 months of GAHT. Both studies evaluated the correlation between bone parameters and serum estradiol concentration, taken as a mean of values performed at months 3 and 12. Wiepjes et al. reported changes in areal BMD in 231 transgender individuals and found that absolute BMD increased at the lumbar spine (LS) (+3.67%, 95% confidence interval [CI]: 3.20 to 4.13, p<0.001), total hip (TH) (+0.97%, 95% CI: 0.62 to 1.31, p<0.001), and femoral neck (FN) (+1.86%, 95% CI: 1.41 to 2.31, p<0.001) for 12 months.³⁰

Estradiol assays differed between sites so cohorts from Amsterdam and Ghent were reported separately. In Amsterdam, serum estradiol concentration was found to correlate with LS, TH, and FN BMD change, whereas in Ghent serum estradiol concentration correlated with LS BMD change but not TH or FN (Table 1).

Vlot et al. evaluated changes in markers of bone metabolism in 121 transgender individuals newly commencing GAHT.³³ They found that alkaline phosphatase decreased in 19% (95% CI: −21 to −16), C-terminal telopeptide of type 1 collagen (CTx) decreased in 11% (95% CI: −18 to −4), and sclerostin decreased in 8% (95% CI: −13 to −4) of individuals after 12 months of GAHT (Table 1). Serum estradiol concentration was reported in quartiles (first quartile 115 pmol/L, second quartile 192 pmol/L, third quartile 280 pmol/L, and fourth quartile 527 pmol/L) in this analysis. The lowest estradiol quartile showed a slight increase in sclerostin concentration, whereas sclerostin decreased in the other quartiles. There was no association between the other markers of bone metabolism and serum estradiol concentration. This could imply that the serum estradiol concentrations achieved in the lowest quartile could be too low to result in reduced bone turnover.

Two retrospective cohort studies from the Amsterdam Cohort of Gender Dysphoria (ACOG) have aimed to establish the long-term changes in areal BMD and fracture risk in transgender individuals. Wiepjes et al. reported changes in absolute BMD and Z-score as measured by DXA in 711 transgender individuals up to 10 years after commencement of GAHT.³⁴ The mean sex steroid concentrations in this study were established by averaging results from laboratory measurements after 1, 2, 5, and 10 years of GAHT and reported in tertiles. Only individuals in the third tertile achieved serum estradiol concentrations recommended in Endocrine Society Clinical Practice Guidelines.

After 10 years of GAHT, LS BMD was not different from baseline (+0.006 g/cm² [−0.005 to +0.017 g/cm²]). However, in the cohort that achieved serum estradiol concentrations in the highest estradiol tertile (mean 442 pmol/L), an increase in LS BMD was observed (+0.044 g/cm² [+0.025 to +0.063 g/cm²]), whereas there was a decrease in those in the lowest tertile (mean 118 pmol/L: −0.026 g/cm² [−0.044 to −0.009 g/cm²]). BMD was stable in the second tertile (mean 238 pmol/L: +0.002 g/cm² [−0.016 to +0.0021 g/cm²]). There was no difference in the change in LS BMD between individuals with suppressed versus not suppressed testosterone.

Wiepjes et al. also established the fracture incidence of transgender individuals treated with long-term GAHT.³⁵ Their study included 2023 transgender individuals; 1089<50 years of age treated with GAHT for a median of 8 years and 934 individuals age 50 years or older treated with GAHT for a median of 19 years. In total 2.4% of younger and 4.4% of older transgender women sustained a fracture. It should be noted that fracture data were obtained from a database of emergency room presentations and may not represent minimal trauma fractures. Compared with age-matched cisgender men, older transgender women had a higher prevalence of fracture (odds ratio [OR]=1.90, 95% CI: 1.32–2.74). Laboratory measurements were available in 66% of individuals. On univariable analyses, no association was found between serum estradiol concentration and fracture risk (per 10 pmol/L: OR=0.99, 95% CI: 0.97–1.02).

Two studies have established volumetric BMD using peripheral quantitative computed tomography (pQCT) in transgender individuals. Serum estradiol concentration at 12 months was not associated with BMD or pQCT parameters in a cohort of 49 transgender individuals.²⁰ Similarly, serum estradiol concentration was not associated with pQCT parameters in a cross-sectional analysis of 23 transgender individuals at least 3 years postgender-affirming surgery.³⁶ Both studies are limited by small numbers of participants and estradiol concentration measured at one timepoint.

In summary, higher serum estradiol concentrations have been associated with higher areal BMD in some studies of transgender individuals. Based on current data, if serum estradiol concentration is maintained >200–250 pmol/L (54–68 pg/mL), serum bone turnover markers are reduced and LS BMD remains stable. Small studies have not found associations between pQCT parameters and serum estradiol concentration.

Risks of escalating estradiol dose/concentration

The risks of escalating estradiol dose to achieve higher serum estradiol concentrations must be weighed against potential adverse events. Higher oral estradiol doses (defined as >1 mg estradiol,^4,37 or >2 mg estradiol or 0.625 mg conjugated equine estrogens³⁸) have been associated with an increased risk of venous thromboembolism with menopausal hormone therapy. However, this has not been demonstrated with high-dose (>50 mcg/24 h) transdermal preparations^4,37,38 apart from one nested case–control study in which there was an increased risk of stroke in women treated with transdermal estradiol >50 mcg/24 h compared with low-dose estradiol.³⁹

Higher oral estradiol dose has also been associated with a higher risk of cholelithiasis in postmenopausal women,⁴⁰ but there have not been reports of an increased risk of cholelithiasis in transgender individuals undergoing feminizing hormone therapy. It is also important to acknowledge that the literature from which this is derived is based on doses often much lower than those administered to transgender individuals, and include studies involving conjugated equine estrogens that are not recommended for GAHT.¹

Higher endogenous serum estradiol concentrations have also been associated with an increased risk of peripheral arterial disease in men⁴¹ and breast cancer in postmenopausal women.⁴² One cohort study of elderly men found an association between higher serum estradiol concentrations and cerebrovascular disease in men,⁴³ but no association was found in a subsequent meta-analysis.⁴⁴ Importantly, higher endogenous serum estradiol concentrations have been associated with lower coronary artery calcium score in postmenopausal women.⁴⁵

Contribution from testosterone suppression

Current guidelines recommend increasing serum estradiol and decreasing serum testosterone into the respective reference ranges of premenopausal women. It is, therefore, difficult to dissect the relative contributions of testosterone suppression from the increases in serum estradiol concentration. Published prospective studies from the ENIGI cohort prescribe cyproterone acetate 25–50 mg daily as antiandrogen therapy,⁴⁶ which results in testosterone suppression in the majority of individuals.⁵ Lower doses of cyproterone acetate result in adequate testosterone suppression.⁴⁷ Importantly, more adequate suppression of serum testosterone concentration has been associated with enhanced breast development as measured by Tanner stage.¹⁰ Similarly, greater changes in body composition are seen in transgender adolescents treated with GnRHa monotherapy than after subsequent commencement of estradiol.²⁶

Limitations of current evidence

The current literature is limited by a small number of observational trials. These studies have based the estradiol concentration from the average of a limited number of serum estradiol concentration measurements with variability in the assay used. This is important given the imprecision of the estradiol immunoassay, particularly at low estradiol concentrations.⁴⁸ It should also be noted that only individuals in the highest quartile or tertile of serum estradiol concentration in the prospective studies achieved concentrations in the range recommended in the Endocrine Society guidelines.¹ The relative contribution of serum estradiol concentration on physical feminization in transgender adults is unknown. Similarly, given that most individuals have a suppressed serum testosterone concentration, the current evidence cannot distinguish the relative contributions of suppressed testosterone and serum estradiol concentration achieved.

Future directions

There is a need for prospective trials evaluating different serum estradiol concentration targets with clinical and radiological features of feminization in larger cohorts of transgender individuals. Ideally, serum estradiol concentration should be measured through liquid chromatography–mass spectrometry at more frequent timepoints. Although sex steroid concentrations have not been found to correlate with sexual desire⁴⁹ or anger⁵⁰ in transgender individuals undergoing feminizing hormone therapy, the influence of serum estradiol concentration on other end-points such as psychological distress or gender dysphoria should also be considered.

Conclusion

Limited uncontrolled prospective evidence has not found that higher serum estradiol concentrations with adequate testosterone suppression enhances breast development or produces more feminine changes to body composition in transgender adults. However, higher serum estradiol concentrations have been associated with higher areal BMD and could be considered in individuals with low bone mass. The precise serum estradiol concentration that results in adequate feminization without increasing the risk of complications remains unknown. Prospective studies with various serum estradiol concentration targets and clinical features of feminization are required.

Abbreviations Used

ACOG

Amsterdam Cohort of Gender Dysphoria

ALP

alkaline phosphatase

BMD

bone mineral density

CI

confidence interval

CTx

C-terminal telopeptide of type 1 collagen

ENIGI

European Network for the Investigation of Gender Incongruence

DXA

dual-energy x-ray absorptiometry

ENIGI

European Network for the Investigation of Gender Incongruence

FN

femoral neck

GAHT

gender-affirming hormone therapy

GnRHa

gonadotropin-releasing hormone agonists

LS

lumbar spine

OR

odds ratio

P1NP

procollagen type 1 N propeptide

pQCT

peripheral quantitative computed tomography

TH

total hip

WHR

waist-hip ratio

Authors' Contributions

B.J.N. reviewed the literature and drafted and revised the article. A.S.C. provided supervision and revised the article.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

B.J.N. is supported by the Royal Australasian College of Physicians Fellows Research Entry Scholarship. A.S.C. is supported by an Australian Government National Health and Medical Research Council Early Career Fellowship (#1143333) and receives research support from the Viertel Charitable Foundation Clinical Investigator Award, Endocrine Society of Australia Postdoctoral Award, and the Royal Australasian College of Physicians.

Cite this article as: Nolan BJ, Cheung AS (2021) Relationship between serum estradiol concentrations and clinical outcomes in transgender individuals undergoing feminizing hormone therapy: a narrative review, Transgender Health 6:3, 125–131, DOI: 10.1089/trgh.2020.0077.